#  @ogawa_tter OGAWA, Tadashi OGAWA, Tadashi posts on X about ai, dram, $googl, huawei the most. They currently have [-----] followers and [---] posts still getting attention that total [-----] engagements in the last [--] hours. ### Engagements: [-----] [#](/creator/twitter::344636014/interactions)  - [--] Week [------] +45% - [--] Month [------] +61% - [--] Months [-------] -17% - [--] Year [---------] -31% ### Mentions: [--] [#](/creator/twitter::344636014/posts_active)  - [--] Week [--] +120% - [--] Month [---] -64% - [--] Months [---] +64% - [--] Year [-----] +16% ### Followers: [-----] [#](/creator/twitter::344636014/followers)  - [--] Week [-----] -0.07% - [--] Month [-----] +0.33% - [--] Months [-----] +2.90% - [--] Year [-----] +9.40% ### CreatorRank: [---------] [#](/creator/twitter::344636014/influencer_rank)  ### Social Influence **Social category influence** [technology brands](/list/technology-brands) 54.84% [stocks](/list/stocks) 47.31% [currencies](/list/currencies) 11.29% [finance](/list/finance) 8.06% [countries](/list/countries) 5.91% [automotive brands](/list/automotive-brands) 2.15% [travel destinations](/list/travel-destinations) 1.61% [events](/list/events) 1.61% [cryptocurrencies](/list/cryptocurrencies) 1.61% [vc firms](/list/vc-firms) 0.54% **Social topic influence** [ai](/topic/ai) 24.19%, [dram](/topic/dram) #178, [$googl](/topic/$googl) 9.68%, [huawei](/topic/huawei) 8.6%, [dec](/topic/dec) #270, [inference](/topic/inference) 6.99%, [systems](/topic/systems) 5.91%, [ceo](/topic/ceo) 5.38%, [core](/topic/core) 4.84%, [ibm](/topic/ibm) 4.3% **Top accounts mentioned or mentioned by** [@1psemicon](/creator/undefined) [@ogawatter](/creator/undefined) [@onurmutlu](/creator/undefined) [@yyasuda](/creator/undefined) [@semianalysis](/creator/undefined) [@jokereph](/creator/undefined) [@roeschinc](/creator/undefined) [@miyox](/creator/undefined) [@ebiken](/creator/undefined) [@ofiberlady](/creator/undefined) [@50](/creator/undefined) [@inouefumihiro](/creator/undefined) [@suekichiii](/creator/undefined) [@cqcqcqdx](/creator/undefined) [@bcantrill](/creator/undefined) [@instlatx64](/creator/undefined) [@yuya0240](/creator/undefined) [@akigoodspeed](/creator/undefined) [@zenrinofficial](/creator/undefined) [@mak0nakamura](/creator/undefined) **Top assets mentioned** [Alphabet Inc Class A (GOOGL)](/topic/$googl) [IBM (IBM)](/topic/ibm) [Microsoft Corp. (MSFT)](/topic/microsoft) [IonQ, Inc. (IONQ)](/topic/$ionq) [Tesla, Inc. (TSLA)](/topic/tesla) [Exxon Mobil (XOM)](/topic/exxon-mobil) [SkyWater Technology, Inc. (SKYT)](/topic/$skyt) [GlobalFoundries (GFS)](/topic/globalfoundries) ### Top Social Posts Top posts by engagements in the last [--] hours "= HUAWEI CONNECT [----] Sep [--] Ascend 950PR (Prefill & Recommend) [----] Q1 950DT (Decode & Training) [----] Q4 [--] PF (FP8)/2 PF (FP4) FP32/HF32/FP16/BF16/FP8/MXFP8/HiF8/MXFP4 Proprietary HBM NVIDIA Rubin CPX Disaggregated https://x.com/ogawa_tter/status/1966342884363497979 https://www.ithome.com/0/883/839.htm = Another Giant Leap: The Rubin CPX Specialized Accelerator & Rack Sep [--] [----] @SemiAnalysis_ https://t.co/RkmMlYxAfV The Giant Leap: Disaggregated Serving lan Buck Sep [--] https://t.co/nFfaV9zR5e Bill Dally Aug [--] https://t.co/ksqVYCR6A5 https://t.co/ssxBrbz3WG Disaggregated" [X Link](https://x.com/ogawa_tter/status/1968787152977674510) 2025-09-18T21:20Z [----] followers, [----] engagements "= Research Progress of High-Degree OXC Technologies Invited Jul [--] [----] Huawei OptiXtrans DC808 Sep [--] [----] Special Prize at Interop Tokyo [----] Jun [--] Beyond CLOS https://x.com/ogawa_tter/status/1976552928443035975 https://e.huawei.com/en/news/2025/solutions/enterprise-optical-network/data-center-all-optical-switch https://e.huawei.com/cn/news/2024/industries/grid/all-optical-switch-optixtrans-dc808 https://www.opticsjournal.net/Articles/OJbc1ca6bc13ed3246/Abstract State of the Opt Components Industry Omdia ECOC [----] Market Focus https://t.co/BRNpAGZdha = Beyond CLOS Futurewei ECOC 2025" [X Link](https://x.com/ogawa_tter/status/1985550170873122966) 2025-11-04T03:30Z [----] followers, [----] engagements "= "IonQ to Acquire SkyWater Technology Creating the Only Vertically Integrated Full-Stack Quantum Platform Company" Jan [--] [----] $1.8B ($35.00/share) SkyWater Jan [--] Strong momentum in quantum computing. https://s27.q4cdn.com/656702482/files/doc_events/2026/Jan/14/SKYT-IR-Presentation-for-NGC-Jan-2026.pdf https://s28.q4cdn.com/828571518/files/doc_downloads/2026/01/26/IonQ-to-Acquire-SkyWater-Technology-2026-01-26-vFF.pdf https://investors.ionq.com/news/news-details/2026/IonQ-to-Acquire-SkyWater-Technology-Creating-the-Only-Vertically-Integrated-Full-Stack-Quantum-Platform-Company/default.aspx" [X Link](https://x.com/ogawa_tter/status/2015790518471651385) 2026-01-26T14:14Z [----] followers, [---] engagements "= "Tower Semiconductor Teams with NVIDIA to Advance AI Infrastructure with 1.6T Data Center Optical Modules" Feb [--] [----] Silicon photonics for 1.6T data center optical modules designed for NVIDIA networking protocols SiPho https://towersemi.com/technology/rf-and-hpa/silicon-photonics-rf/ https://towersemi.com/2026/02/05/02052026/ https://towersemi.com/technology/rf-and-hpa/silicon-photonics-rf/ https://towersemi.com/2026/02/05/02052026/" [X Link](https://x.com/ogawa_tter/status/2019983493787447484) 2026-02-07T03:56Z [----] followers, [---] engagements "= Select Committee on the CCP REPORT: American VC Firms Investing Billions into PRC Companies Fueling the CCP's Military Surveillance State and Uyghur Genocide Feb [--] [----] GGV Capital GSR Ventures Qualcomm Ventures Sequoia Capital (China) Walden Inter https://selectcommitteeontheccp.house.gov/media/press-releases/committee-report-american-vc-firms-investing-billions-prc-companies-fueling https://selectcommitteeontheccp.house.gov/media/press-releases/committee-report-american-vc-firms-investing-billions-prc-companies-fueling" [X Link](https://x.com/ogawa_tter/status/1755886193911500855) 2024-02-09T09:27Z [----] followers, 20.3K engagements "= William James Dally [----] Benjamin Franklin Medal in Computer & Cognitive Science Merrimac SC03 Brook for GPUs SIGGRAPH [----] Chris Malachowsky NVIDIA EDPS [----] Stream Processors ACM Queu Mar [----] https://x.com/ogawa_tter/status/1611750717865463808 https://x.com/ogawa_tter/status/1733150189450412442 https://fi.edu/en/awards/laureates/william-james-dally = Stream Processors Mar [----] https://t.co/SG9FO9aHAm SIGRAPH [----] CUDA Jun [----] https://t.co/wNi9fSqDDI Scalable Parallel Programming with CUDA Mar/Apr [----] https://t.co/9PapnzQYMs The GPU Computing Era Micro Mar/Apr 2010" [X Link](https://x.com/ogawa_tter/status/1987143958276059280) 2025-11-08T13:03Z [----] followers, [----] engagements ""Government University & Industry Cooperation: The NVIDIA Story" Bill Dally NSB Meeting Jul [--] [----] D. Kirk SC19 IAA WS [----] = Benjamin Franklin Medal https://x.com/ogawa_tter/status/1987143958276059280 https://x.com/ogawa_tter/status/1599768199759040513 https://x.com/ogawa_tter/status/1205627606533464064 https://nsf-gov-resources.nsf.gov/files/SPEAKER_2_BDally_panel_slides.pdf https://www.youtube.com/live/9Os9sAAcsjMt=842s = William James Dally [----] Benjamin Franklin Medal in Computer & Cognitive Science https://t.co/OtFWa4CqZP https://t.co/WucQGgalko Merrimac SC03 Brook for GPUs SIGGRAPH" [X Link](https://x.com/ogawa_tter/status/1990422799929004304) 2025-11-17T14:12Z [----] followers, [----] engagements "= "Addressing Memory Contention in Modern DRAM-Based Systems" M. Patel Rutgers SAFARI Conf Dec [--] [----] PhD Thesis [----] @_onurmutlu_ R/W in parallel POWER9=Power10 Google Appl https://x.com/ogawa_tter/status/1896529002199970293 https://x.com/ogawa_tter/status/1879535899350085707 https://events.safari.ethz.ch/SafariConference25/lib/exe/fetch.phpmedia=minesh-safari-conference-contention-afterlecture.pptx https://www.youtube.com/live/QtS7v8aVGLQt=713s = Google Systems & Methods for HBM with Unidirectional Data Flow (P: Aug 2023) https://t.co/aCwOci14fs Concurrent Read & Write in parallel HBM4" [X Link](https://x.com/ogawa_tter/status/2002744203726700723) 2025-12-21T14:13Z [----] followers, [----] engagements "= "Efficient Kernel Mapping and Comprehensive System Evaluation of LLM Acceleration on a CGLA" T. Ando . Y. Nakashima ( ) NAIST (LENZO) IEEE Access Dec [--] [----] LENZO [-------] [----] IMAX (CGLA) NAIST https://x.com/ogawa_tter/status/1987877170291876102 https://ieeexplore.ieee.org/document/11264489 CGLANVIDIA2025827 https://t.co/Qz7I42l8BG LENZO NVIDIA = [----] https://t.co/EAoQ5QlZJW CGRA vs CGLA Jul [--] VPP500 ICS '95 IMAX (CGLA) NAIST https://t.co/DpUKXPUyIr https://t.co/Y6hOwEf3CO https://x.com/ogawa_tter/status/1987877170291876102 https://ieeexplore.ieee.org/document/11264489 CGLANVIDIA2025827" [X Link](https://x.com/ogawa_tter/status/2007831406567903258) 2026-01-04T15:08Z [----] followers, [----] engagements "NVIDIA CGLA202611 "Architecture Leads System Design" Y. Nakashima NAIST/LENZO Invited Nov [--] [----] = IEEE Access Dec [--] [----] FPGA Prototype https://x.com/ogawa_tter/status/2007831406567903258 https://www.youtube.com/watchv=gV90-tfVTbg https://www.youtube.com/watchv=1LBNnC37CKI = "Efficient Kernel Mapping and Comprehensive System Evaluation of LLM Acceleration on a CGLA" T. Ando . Y. Nakashima ( ) NAIST (LENZO) IEEE Access Dec [--] [----] https://t.co/qflvzQnf3H LENZO [-------] https://t.co/79Ogr7aMHV [----] IMAX (CGLA) NAIST https://t.co/EWlm1GI8i2 https://x.com/ogawa_tter/status/2007831406567903258" [X Link](https://x.com/ogawa_tter/status/2007997326821077432) 2026-01-05T02:07Z [----] followers, [----] engagements "= Advances and Future Trends in Photolithography & Photoresist Materials Invited Apr [----] [--] pp [---] ref Research Progress & Prospects of Extreme Ultraviolet Photoresists Sep [----] May [----] https://x.com/ogawa_tter/status/1653573723746881536 https://www.opticsjournal.net/Articles/OJf78e7687596f5910/Abstract https://www.opticsjournal.net/Articles/OJ762ec4ae17d37064/Abstract = [-----] EUV Lithography Source https://t.co/M8succaWMF Stage Control Methods https://t.co/7a3TXtSyOx EUV Photoresists https://t.co/ON84ttXsLy Computational Lithography https://t.co/g8bGJhxFTm Advanced Photolithography" [X Link](https://x.com/ogawa_tter/status/2011420127544381834) 2026-01-14T12:48Z [----] followers, [---] engagements "= GlobalFoundriestoAcquireSynopsysProcessor IPSolutions Business ExpandingCapabilitiesto Accelerate Physical AIApplications Jan [--] [----] ARC-V ARC-Classic ARCVPX-DSPandARCNPX NPU ASIP tools GF: Physical AI Dec [--] https://investors.gf.com/events/event-details/gf-business-webinar-series-gf-enabling-evolution-edge-physical-ai https://gf.com/gf-press-release/globalfoundries-to-acquire-synopsys-processor-ip-solutions-business/ https://investors.gf.com/events/event-details/gf-business-webinar-series-gf-enabling-evolution-edge-physical-ai" [X Link](https://x.com/ogawa_tter/status/2011630756267987059) 2026-01-15T02:45Z [----] followers, [---] engagements "= "SiFive to Power Next-Gen #RISCV AI Data Centers with NVIDIA NVLink Fusion" Jan [--] [----] NVLink Fusion HotI [----] https://x.com/ogawa_tter/status/1962326039646097646 https://www.nvidia.com/en-us/data-center/nvlink-fusion/ https://www.theregister.com/2026/01/15/sifive_nvidia_nvlink/ https://www.sifive.com/press/sifive-nvidia-nvlinkfusion-datacenter = Building Custom AI Infrastructure w/ NVLink Fusion K Geeyarpuram Invited HotI [----] Aug [--] https://t.co/nazZi87SOl https://t.co/lMJHANLJag NVLink & NVLink Fusion Aug [--] https://t.co/uLmE6tSDkH "FugakuNEXT" Aug [--] https://t.co/FJdsUNdhIc Fujitsu" [X Link](https://x.com/ogawa_tter/status/2011999403780030685) 2026-01-16T03:10Z [----] followers, [---] engagements "= "Synopsys Enters Definitive Agreement with GlobalFoundries For Sale of Processor IP Solutions Business" Jan [--] [----] Synopsys Processor Solutions ARC-V (#RISCV) etc GF Physical AI https://x.com/ogawa_tter/status/2011630756267987059 https://www.synopsys.com/dw/doc.php/ds/cc/dw-processor-solutions.pdf https://www.synopsys.com/designware-ip/processor-solutions.html https://investor.synopsys.com/news/news-details/2026/Synopsys-Enters-Definitive-Agreement-with-GlobalFoundries-For-Sale-of-Processor-IP-Solutions-Business/default.aspx = GlobalFoundriestoAcquireSynopsysProcessor IPSolutions Business" [X Link](https://x.com/ogawa_tter/status/2012050607474163890) 2026-01-16T06:33Z [----] followers, [---] engagements "= IBM Buffered Server Memory POWER9 Power10: [----] GB/s OMI DDR4 Power11 Jul [----] [----] GB/s OMI DDR5 [--] TB [----] GB/s/Skt Credo OmniConnect Core AXI over 112G VSR = Weaver: LPDDR Positron Asimov N3P ASIC https://x.com/ogawa_tter/status/2012345610956108247 https://credosemi.com/products/omniconnect/ https://x.com/ogawa_tter/status/1950024476684419076 = Breaking the Memory Wall: Scaling AI Inference with Innovative Memory Fanout Arch On-Demand (Nov [--] 2025) https://t.co/nHCZ0RuB8N Credo Nov [--] https://t.co/468eqr6Ol8 OmniConnect 112G VSR Weaver Memory Fanout Gearbox M. Agrawal CEO Positron AI Dec" [X Link](https://x.com/ogawa_tter/status/2012383715750711513) 2026-01-17T04:37Z [----] followers, [----] engagements "= Google Appls (PCT: Mar 2025) Memory & Compute 3DIC Integration without TSV Thermal Management for Integrated Memory & Compute Circuits X. Ma & D. Patterson arXiv Jan [--] (8) [----] Processing-Near-Memory https://x.com/ogawa_tter/status/2010264445755548026 https://patents.google.com/patent/WO2025189071A1/en https://patents.google.com/patent/WO2025208140A1/en = Xiaoyu Ma David Patterson. Google arXiv Jan [--] [----] https://t.co/lvUhAzkOqo "Tree-based Network Architecture for Accelerating Machine Learning Collective Operations" Nov [----] (Jan 2024) https://t.co/R5vQkMJY1j LLM with 3D-DRAM Chiplets (P:" [X Link](https://x.com/ogawa_tter/status/2012425522119119280) 2026-01-17T07:23Z [----] followers, [---] engagements "= "Critical Minerals on the U.S. Outer Continental Shelf: The Bureau of Ocean Energy Managements Role and Issues for Congress" Jan [--] [----] CRS EO [-----] Apr [----] Mineral Commodity Summaries [----] USGS https://x.com/ogawa_tter/status/1886958547482894345 https://www.whitehouse.gov/presidential-actions/2025/04/unleashing-americas-offshore-critical-minerals-and-resources/ https://www.congress.gov/crs-product/R48302 = Decision to Implement Export Control on Items Related to W Te Bi Mo & In Feb [--] [----] https://t.co/x4BRohlGEq USGS Mineral Commodity Summaries [----] Jan [--] https://t.co/9VORlON5Wy Ga&Ge" [X Link](https://x.com/ogawa_tter/status/2014657222849265876) 2026-01-23T11:11Z [----] followers, [---] engagements "Water Bottle Memory (2T Gain Cell Memory) Shuhan Liu Stanford 2nd MemoryDAX/DAM Winter WS Jan [--] [----] IEDM25 w/ TSMC NVIDIA Oxide Semicon for Memory Review Jan [--] = 3T GC-RAM https://x.com/ogawa_tter/status/2014664852078526734 https://x.com/ogawa_tter/status/2010608213436289284 https://x.com/ogawa_tter/status/1996939609570529777 https://dam.stanford.edu/assets/2026-workshop/dam-ww26-liu.pdf = "Design Optimization and Verification of Embedded Gain Cell RAMs" PhD Thesis [----] https://t.co/sfAOVXgrCM 3T GC-RAM in [--] nm CMOS [--] nm FDSOI [--] nm FinFET https://t.co/Gi0kaVQJqD RAAAM Memory" [X Link](https://x.com/ogawa_tter/status/2015750464906956867) 2026-01-26T11:35Z [----] followers, [----] engagements "= "A Device and Method for Reducing Memory Access Delay of 3D Stacked DRAM" NUDT (CN: Nov 2024) TSV & the Row buffers and Physical circuits form a high-speed data exchange bridge Stacked DRAM w/o TSV Huawei Appl (CN: Apr 2024) https://x.com/ogawa_tter/status/2004922428020183375 https://patents.google.com/patent/CN119938553A/en = "Semiconductor Packaging Structure and Electronic Device" Huawei Appl (CN: Apr 2024) https://t.co/iZuCutuRbj Stacked DRAM (400) without TSV Appls https://t.co/PrbrXhve4d DRAM (CN: Nov 2021) Stacked DRAM (CN: Dec 2021) Proprietary HBM: HiBL / HiZQ HUAWEI CONNECT 2025" [X Link](https://x.com/ogawa_tter/status/2017575761268613552) 2026-01-31T12:28Z [----] followers, [---] engagements "= "Thermal Management Challenges in 2.5D and 3D Chiplet Integration: A Review on ArchitectureCooling Co-Design" Eng Dec [--] [----] [--] pp [---] ref HB & HBM Jan [--] [----] HP & NVIDIA Microfluidic Cooling https://x.com/ogawa_tter/status/1971941691994300867 https://x.com/ogawa_tter/status/2017233675826675880 https://www.mdpi.com/2673-4117/6/12/373 = Trump administration cancels $45 million Oregon State grant for Microfluidics research (CorMic) May [--] [----] https://t.co/oQUQSssdKj OSU https://t.co/sxwqj4ME2i CHIPS & Science Act EDA Oct [----] https://t.co/7YerhcPjdu HP & NVIDIA ARPA-E COOLERCHIPS" [X Link](https://x.com/ogawa_tter/status/2018333361459347831) 2026-02-02T14:39Z [----] followers, [---] engagements "@1p_semicon @InoueFumihiro IEEE Hybrid Bonding Symposium Jan [--] [----] John H Lau (Unimicron) HB ()/ https://attend.ieee.org/hbs/page_id=779 https://x.com/ogawa_tter/status/2018165677656539340 = "From Interface Chemistry to Bond Wave: Unraveling SiCN/Cu Bonding Mechanisms" @InoueFumihiro YNU IEEE Hybrid Bonding Symp Jan [--] [----] (38:30) https://t.co/4PBaSUj5qQ https://t.co/glioDMnq5x https://t.co/9LvQUQQjhX Recent Advances & Trends Jan [--] https://t.co/XZhboN2mvn https://t.co/Z8BJVv6IVs https://attend.ieee.org/hbs/page_id=779 https://x.com/ogawa_tter/status/2018165677656539340 = "From Interface" [X Link](https://x.com/ogawa_tter/status/2018526227678757229) 2026-02-03T03:25Z [----] followers, [----] engagements "= [----] COOLERCHIPS Annual Review Project OMNICOOL Ali Heydari NVIDIA Nov [--] Ali Heydari Microfluidic Silicon Cold Plate HP & NVIDIA Nov [--] Trump Admin cancels $45M May [--] [----] https://x.com/ogawa_tter/status/2018333361459347831 https://arpa-e.energy.gov/sites/default/files/2026-01/Day2_03a_HPI_Cumbie.pdf https://x.com/ogawa_tter/status/1884921947131420948 https://arpa-e.energy.gov/sites/default/files/2026-01/Day1_04a_NVIDIA_Manaserh.pdf = "Thermal Management Challenges in 2.5D and 3D Chiplet Integration: A Review on ArchitectureCooling Co-Design" Eng Dec [--] [----] https://t.co/viXciJft0G [--] pp" [X Link](https://x.com/ogawa_tter/status/2018640263540068641) 2026-02-03T10:58Z [----] followers, [---] engagements "= "Package Architectures having Vertically Stacked Dies for High Capacity Memory" Intel (Dec 2023) High Capacity & High Bandwidth Memory while reducing the complexity of the package relative to conventional approaches J. Fryman = https://x.com/ogawa_tter/status/1704435232189882434 https://patents.google.com/patent/US20250201793A1/en = Intel. DARPA ERI Summit [----] Aug [--] Pat Gelsinger Opening https://t.co/pNO4dMbmXD https://t.co/LBK1e7iGdt Univ Shuttle https://t.co/HEdgpQs151 by Complex 3D Microsystems J. Fryman Intel Fellow https://t.co/HKPydLXJpo https://t.co/FWfmKIcxAR Sep 19" [X Link](https://x.com/ogawa_tter/status/2019040243077570752) 2026-02-04T13:28Z [----] followers, [---] engagements "@1p_semicon ( Advanced Memory Technology (AMT) DoE NNSA Intel AMT Jan [--] AMT (Sandia) https://x.com/ogawa_tter/status/1723159963995083206 = Projects for the Advanced Memory Technologies and Vendor R&D Product ASC Aug [--] [----] https://t.co/pRtfKJ1Jx5 R&D (SNL) Cerebras Oct [--] [----] https://t.co/3VNqEJY4JM Intel Dec [--] https://t.co/AdoTiDEWNK AMD Jan [--] [----] https://t.co/3kC5Vwg7W3 NVIDIA https://t.co/LrPEAPdKcL https://t.co/ovSx7e2BEp https://x.com/ogawa_tter/status/1723159963995083206 = Projects for the Advanced Memory Technologies and Vendor R&D Product ASC Aug [--] [----] https://t.co/pRtfKJ1Jx5 R&D" [X Link](https://x.com/ogawa_tter/status/2019219696164434289) 2026-02-05T01:21Z [----] followers, [---] engagements "= Positron AI Raises $230M Series B Feb [--] [----] Asimov HWed SoftMax RMSNorm RoPE SwiGLU Multiple on-chip ARMv9 Scale-Out Interconnect Credo OmniConnect Core AXI over 112G VSR = Weaver: LPDDR *) Credo & Positron AI On-Demand https://x.com/ogawa_tter/status/2012383715750711513 https://www.businesswire.com/news/home/20260204250472/en/Positron-AI-Raises-%24230-Million-Series-B-at-Over-%241-Billion-Valuation-to-Scale-Energy-Efficient-AI-Inference = IBM Buffered Server Memory POWER9 Power10: [----] GB/s OMI DDR4 Power11 Jul [----] https://t.co/cWKGugCfg6 [----] GB/s OMI DDR5 [--] TB [----] GB/s/Skt Credo" [X Link](https://x.com/ogawa_tter/status/2019260765514133790) 2026-02-05T04:04Z [----] followers, [---] engagements "Vertically Stacked Dies with a Cooling Microchannel Intel Appl (Sep 2023) = Thermal Mgmnt Challenges in 3D Review Dec [----] J. Fryman Intel ERI Summit [----] SAIMEMORY Feb [--] [----] https://x.com/ogawa_tter/status/2019249282877125075 https://x.com/ogawa_tter/status/1704435232189882434 https://x.com/ogawa_tter/status/2018333361459347831 https://patents.google.com/patent/US20250079263A1/en "SAIMEMORY"202623 https://t.co/O2FnHYEy8C [----] Eng https://t.co/YgWteRQQZV = https://t.co/B6ghH06aS2 Intel Feb [--] EE Times [--] AMT (NNSA/Sandia) Jan [--] Intel Appl (Dec 2023) https://t.co/hhekUv4KS8" [X Link](https://x.com/ogawa_tter/status/2019367145742954949) 2026-02-05T11:07Z [----] followers, [----] engagements "SAIMEMORY [------] [--] [--] NEXT S. Matsuoka Oct [----] = Fujitsu (Intel Appl) MONAKA-X+ZAM https://x.com/ogawa_tter/status/2019249282877125075 https://x.com/ogawa_tter/status/2018518028699464016 https://x.com/ogawa_tter/status/1983821426307297402 https://xtech.nikkei.com/atcl/nxt/column/18/00001/11457/P=2 https://txbiz.tv-tokyo.co.jp/txn/news_txn/post_334408 "SAIMEMORY"202623 https://t.co/O2FnHYEy8C [----] Eng https://t.co/YgWteRQQZV = https://t.co/B6ghH06aS2 Intel Feb [--] EE Times [--] AMT (NNSA/Sandia) Jan [--] Intel Appl (Dec 2023) https://t.co/hhekUv4KS8" [X Link](https://x.com/ogawa_tter/status/2019747974952145110) 2026-02-06T12:20Z [----] followers, [----] engagements "= Vertically Stacked Dies Intel Appls Voltage (Sep 2023) (Dec 2023) Power Delivery Cooling (Sep 2023) SAIMEMORY Feb [--] [----] High Capacity Memory (Dec 2023) https://x.com/ogawa_tter/status/2019367145742954949 https://patents.google.com/patent/US20250201768A1/en https://patents.google.com/patent/US20250201766A1/en https://patents.google.com/patent/US20250210587A1/en https://patents.google.com/patent/US20250079398A1/en Vertically Stacked Dies with a Cooling Microchannel Intel Appl (Sep 2023) https://t.co/qnCwLvlvbm = Thermal Mgmnt Challenges in 3D Review Dec [----] https://t.co/xUWfuMTxOM J. Fryman" [X Link](https://x.com/ogawa_tter/status/2019783332700029334) 2026-02-06T14:40Z [----] followers, [---] engagements "= Honda and Mythic: Joint Development of 100x Energy-Efficient Analog AI Chip for Next-Gen Vehicles Feb [--] [----] An automotive-grade AI SOC for deployment in Hondas next-gen SDVs by the late 2020s/early 2030s raised $125M Dec [--] https://mythic.ai/whats-new/mythic-to-challenge-ais-gpu-pantheon-with-100x-energy-advantage-and-oversubscribed-125m-raise/ https://mythic.ai/whats-new/honda-and-mythic-announce-joint-development-of-100x-energy-efficient-analog-ai-chip-for-next-generation-vehicles/" [X Link](https://x.com/ogawa_tter/status/2020062999080497596) 2026-02-07T09:12Z [----] followers, [---] engagements "= "Storage and Device" Huawei Appl (CN: Jun 2023) Stacked Storage Medium Wafer & Mmedium Controller Wafer Huawei Appls DRAM (CN: Nov 2021) "Wafer-on-Wafer Formed Memory and Logic" Micron (P: Aug 2021) https://x.com/ogawa_tter/status/1829500736302129505 https://x.com/ogawa_tter/status/1970494560678879673 https://patents.google.com/patent/US20260017196A1/en = Micron: Patent Applications (P: Aug 2021) "Wafer-on-Wafer Formed Memory and Logic" https://t.co/T2QK1beHb9 Logic devices can be AI accelerators can provide a greater bandwidth Testing https://t.co/9pNwTixzgN Signal Routing Input/Output" [X Link](https://x.com/ogawa_tter/status/2020361159770312774) 2026-02-08T04:56Z [----] followers, [---] engagements "= "Reviving Moore's Law with an EUV FEL" Nicholas Kelez CEO & CTO xLight Keynote EUV-FEL WS Jan [--] [----] Signs $150M LoI with the US DoC Dec [--] [----] K-Program: = https://x.com/ogawa_tter/status/2019582917706727722 https://x.com/ogawa_tter/status/1995759750395703519 https://conference-indico.kek.jp/event/350/contributions/7410/attachments/4960/8250/Presentation_Nicholas.pdf = Development of Next-Generation Semiconductor Process Technologies for EUV and BEUV under Japans K Program by JST Keynote 10th EUV-FEL WS Jan [--] [----] https://t.co/Qx6d1mkYQW [--] projects JPY 13.5B [--] yrs" [X Link](https://x.com/ogawa_tter/status/2020691317278535797) 2026-02-09T02:48Z [----] followers, [---] engagements "= Xian UniIC () Initiates IPO Preparations A Core Force in Domestic Memory Advances into the Capital Market Jan [--] [----] SeDRAM Nov [----] https://www.unisemicon.com/index.phpm=content&c=index&a=show&catid=24&id=181 https://x.com/ogawa_tter/status/1926825927574499677 http://en.unisemicon.com/product/qrsdram.html https://xinsanban.eastmoney.com/Article/NoticeContentid=AN202601061816086745 https://www.eet-china.com/mp/a465353.html = "A 3D Unified Analysis Method (3D-UAM) for Wafer-on-Wafer Stacked Near-Memory Structure" USTC & Xian UniIC Semicon IEEE Trans on VLSI System May [--] [----] (Early)" [X Link](https://x.com/ogawa_tter/status/2020751418228171048) 2026-02-09T06:47Z [----] followers, [---] engagements "= "The Vast Light Source Building the world's most powerful lasers " Y. Honda Accelerator Lab KEK Invited EUV-FEL WS Jan [--] [----] EUV-FEL WS 2024/3/2 Jan [--] xLight K-Program https://x.com/ogawa_tter/status/2019582917706727722 https://x.com/ogawa_tter/status/2020691317278535797 https://x.com/ogawa_tter/status/1801163094103589201 https://conference-indico.kek.jp/event/350/contributions/7404/attachments/4957/8241/Presentation_Honda.pdf = Development of Next-Generation Semiconductor Process Technologies for EUV and BEUV under Japans K Program by JST Keynote 10th EUV-FEL WS Jan [--] 2026" [X Link](https://x.com/ogawa_tter/status/2020788559368560837) 2026-02-09T09:15Z [----] followers, [---] engagements "= "Uni-STC: Unified Sparse Tensor Core" CUPB & NUDT HPCA [----] (PPTX) Sparse kernel: SpMV SpMSpV SpMM & SpGEMM Dataflow: Outer-product plus Segmented dot-product (A group of parallel vector mul. vector to update scalars)/cycle https://www.ssslab.cn/assets/slides/2026-lian-UniSTC.pptx https://www.ssslab.cn/assets/papers/2026-lian-UniSTC.pdf https://www.ssslab.cn/assets/slides/2026-lian-UniSTC.pptx https://www.ssslab.cn/assets/papers/2026-lian-UniSTC.pdf" [X Link](https://x.com/ogawa_tter/status/2021070387585614064) 2026-02-10T03:55Z [----] followers, [---] engagements "= "Characterizing Matrix Multiplication Units across General Parallel Patterns in Scientific Computing" CUPB & BSC PPoPP [----] NVIDIA A100/H200/B200 GEMV/GEMM/SpMV/SpGEMM FFT/Stencil/Reduction/Scan/BFS/PiC https://zenodo.org/records/17725527 https://www.ssslab.cn/assets/slides/2026-lu-Cubie.pdf https://dl.acm.org/doi/10.1145/3774934.3786456 https://zenodo.org/records/17725527 https://www.ssslab.cn/assets/slides/2026-lu-Cubie.pdf https://dl.acm.org/doi/10.1145/3774934.3786456" [X Link](https://x.com/ogawa_tter/status/2021115932815524236) 2026-02-10T06:56Z [----] followers, [---] engagements "= "High Bandwidth Destructive Read embedded Memory" Intel Granted: Jun [----] (Jun 2019) 2019-06-29: Application filed by Intel 2025-09-18 Assignment Owner name: SK HYNIX NAND PRODUCT SOLUTIONS CORP. (DBA SOLIDIGM) CALIFORNIA Effective date: [--------] 🤔 https://x.com/ogawa_tter/status/1599027773552885762 = "High Bandwidth Destructive Read embedded Memory" Intel Granted: Jun [----] (Jun 2019) https://t.co/blqOPqK3xx Low-area-overhead increases to the B/W of destructive read embedded memories (eDRAM) such as for caching https://t.co/lHIOnDWXWP DRAM NVIDIA https://t.co/kEl5eyV6Y0" [X Link](https://x.com/ogawa_tter/status/2021125025185792442) 2026-02-10T07:32Z [----] followers, [---] engagements "RT @suekichiii:" [X Link](https://x.com/ogawa_tter/status/2021200270341046339) 2026-02-10T12:31Z [----] followers, [---] engagements "RT @cqcqcqdx: First Public Demonstration of the Computer Mouse and Graphical User Interface [----] On December [--] [----] in what would late" [X Link](https://x.com/ogawa_tter/status/2021212534955504033) 2026-02-10T13:19Z [----] followers, [---] engagements "= "MXFP: A Metadata-Augmented Microscaling Data Format for Efficient Low-bit Quantization" SJTU & Computing Product Line Huawei ASPLOS [----] Mar [--] For Huawei Ascend 950PR (26 Q1)/950DT (26 Q4) FP8/MXFP8/HIF8: [--] PF MXFP4: [--] PF https://x.com/ogawa_tter/status/1968787152977674510 https://arxiv.org/abs/2601.19213v2 = HUAWEI CONNECT [----] Sep [--] https://t.co/6QXsLjPFU6 Ascend 950PR (Prefill & Recommend) [----] Q1 950DT (Decode & Training) [----] Q4 [--] PF (FP8)/2 PF (FP4) FP32/HF32/FP16/BF16/FP8/MXFP8/HiF8/MXFP4 Proprietary HBM NVIDIA Rubin CPX https://t.co/Hi2mWyQqeT Disaggregated" [X Link](https://x.com/ogawa_tter/status/2021223019721466212) 2026-02-10T14:01Z [----] followers, [---] engagements "RT @bcantrill: Our $200M Series C https://oxide.computer/blog/our-200m-series-c https://oxide.computer/blog/our-200m-series-c" [X Link](https://x.com/ogawa_tter/status/2021235222478463191) 2026-02-10T14:50Z [----] followers, [--] engagements "RT @InstLatX64: My current knowledge of #AMD #EPYC topologies: #Zen #Zen2 #Zen3 #Zen4 #Zen5 #Zen6 #Zen7" [X Link](https://x.com/ogawa_tter/status/2021237357001748603) 2026-02-10T14:58Z [----] followers, [--] engagements "RT @yuya0240:" [X Link](https://x.com/ogawa_tter/status/2021237634706592173) 2026-02-10T14:59Z [----] followers, [---] engagements "RT @aki_goodspeed: ✳" [X Link](https://x.com/ogawa_tter/status/2021240063988732374) 2026-02-10T15:09Z [----] followers, [---] engagements "RT @ZENRIN_official: 281" [X Link](https://x.com/ogawa_tter/status/2021396546781315543) 2026-02-11T01:31Z [----] followers, [----] engagements "RT @Mak0Nakamura: ()NIKE" [X Link](https://x.com/ogawa_tter/status/2021438938213384317) 2026-02-11T04:19Z [----] followers, [---] engagements "= "Full duplex DRAM for Tightly Coupled Compute Die and Memory Die" R. Osborne Intel (Aug 2022) HASWELL IEEE Micro Mar/Apr [----] eDRAM: Full-duplex CXL Full-duplex Google Appl https://x.com/ogawa_tter/status/2002744203726700723 https://x.com/ogawa_tter/status/2003433315152576960 https://pages.cs.wisc.edu/rajwar/papers/ieee_micro_haswell.pdf https://patents.google.com/patent/US20220392519A1/en = "Addressing Memory Contention in Modern DRAM-Based Systems" M. Patel Rutgers SAFARI Conf Dec [--] [----] https://t.co/Al5ItSQcAN https://t.co/bXiQx1A8Q4 PhD Thesis [----] https://t.co/BiauTLdXfy @_onurmutlu_" [X Link](https://x.com/ogawa_tter/status/2021503395916546317) 2026-02-11T08:35Z [----] followers, [---] engagements "RT @haskap1017:" [X Link](https://x.com/ogawa_tter/status/2021517874310132161) 2026-02-11T09:33Z [----] followers, [----] engagements "RT @S_Y_yama:" [X Link](https://x.com/ogawa_tter/status/2021591944925794324) 2026-02-11T14:27Z [----] followers, [---] engagements "RT @taku_photo47: 2" [X Link](https://x.com/ogawa_tter/status/2021600568574165183) 2026-02-11T15:01Z [----] followers, [---] engagements "RT @VRGSKtaQwz7ECBb:" [X Link](https://x.com/ogawa_tter/status/2021610015820263895) 2026-02-11T15:39Z [----] followers, [---] engagements "RT @FelixCLC_: This is my quarterly request to @GoogleAI to officially spec BF16 ideally via OCP. This is my quarterly request to @NVIDI" [X Link](https://x.com/ogawa_tter/status/2021610677438144796) 2026-02-11T15:41Z [----] followers, [--] engagements "RT @ASMLcompany: Our extreme ultraviolet (EUV) lithography systems print features just [--] nm wide orders of magnitude smaller than the le" [X Link](https://x.com/ogawa_tter/status/2021656496338338035) 2026-02-11T18:44Z [----] followers, [---] engagements "RT @__tameiki__: AI" [X Link](https://x.com/ogawa_tter/status/2021761660093104446) 2026-02-12T01:41Z [----] followers, [--] engagements "@yyasuda ( [----] [--] DRAMIntel FugakuNext ECC (RAS) Full duplex DRAM (2/5) https://x.com/ogawa_tter/status/2021503395916546317 = "Full duplex DRAM for Tightly Coupled Compute Die and Memory Die" R. Osborne Intel (Aug 2022) https://t.co/78SS6h6D0e HASWELL IEEE Micro Mar/Apr [----] https://t.co/FnUPZnMuXU eDRAM: Full-duplex CXL Full-duplex https://t.co/kbVDaE63ye https://t.co/viHtbBccqt Google Appl https://t.co/0LGhuw7JOS https://x.com/ogawa_tter/status/2021503395916546317 = "Full duplex DRAM for Tightly Coupled Compute Die and Memory Die" R. Osborne Intel (Aug 2022) https://t.co/78SS6h6D0e" [X Link](https://x.com/ogawa_tter/status/2021764321353904613) 2026-02-12T01:52Z [----] followers, [--] engagements "@yyasuda ( MONAKA-X (MSHR)NoC DoE ECP CPU+HBM (1st [--] 2) (3/5) https://x.com/ogawa_tter/status/1591674872253779968 = "HBM2/3 Evaluation on Many-Core CPUs" Jan [--] [----] https://t.co/HNLIP8PKaH Report Jun [--] [----] https://t.co/1HxW1Dg8it HPCG (XSBench) "higher NoC bandwidth and larger MSHR banks are critical for fully exploiting the performance benefits of HBM3" MSHR https://t.co/W00pkKyrsN https://t.co/Xm9icTpVXN https://x.com/ogawa_tter/status/1591674872253779968 = "HBM2/3 Evaluation on Many-Core CPUs" Jan [--] [----] https://t.co/HNLIP8PKaH Report Jun [--] [----] https://t.co/1HxW1Dg8it HPCG (XSBench)" [X Link](https://x.com/ogawa_tter/status/2021764441373892937) 2026-02-12T01:53Z [----] followers, [--] engagements "= NEDO20251218 [--] () SAIMEMORY202623 FugakuNEXT25 MONAKA-X & Intel ZAM (Full duplex DRAM) + NVIDIA GPU Stacked LLC (GC Memory) https://x.com/ogawa_tter/status/2015416350328656371 https://x.com/ogawa_tter/status/2021503395916546317 https://x.com/ogawa_tter/status/2020398753103093990 = "Three Dimensional Chip and Package Integration with Backside Metallization" NVIDIA (Sep 2024) https://t.co/IgRg0rTB4J https://t.co/ZeNTjatJiT COPA-GPU: Patent Mar [----] (P: Feb 2021) / ACM TACO Face-to-Face Dies Patents (Jan 2020) GC Memory IEDM25 https://t.co/DOugYZznRW https://t.co/yN3I3tz9RQ" [X Link](https://x.com/ogawa_tter/status/2021867035987468600) 2026-02-12T08:40Z [----] followers, [---] engagements "= Mueon Emerges From Stealth with $15.5M Seed to Redefine DCs for the AI Era Sep [--] [----] Cubelets Wilfred Gomes Co-founder & CEO Ponte Vecchio Vertically Stacked Dies Intel Appls https://x.com/ogawa_tter/status/2019367145742954949 https://x.com/ogawa_tter/status/2019783332700029334 https://x.com/ogawa_tter/status/1511475822057582593 https://www.intelcapital.com/mueon-emerges-from-stealth-with-15-5m-seed-to-redefine-data-centers-for-the-ai-era/ Vertically Stacked Dies with a Cooling Microchannel Intel Appl (Sep 2023) https://t.co/qnCwLvlvbm = Thermal Mgmnt Challenges in 3D Review Dec 2025" [X Link](https://x.com/ogawa_tter/status/2021935819871899826) 2026-02-12T13:13Z [----] followers, [---] engagements "@1p_semicon ( Intel "Package Architectures having Vertically Stacked Dies for High Capacity Memory" (202312) "Memory Layers Bonded to Logic Layers with Inclination" [-----] DRAM/SRAM ( https://x.com/ogawa_tter/status/2022156337065734249 = Vertically Stacked Dies Intel Appls "Memory Layers Bonded to Logic Layers with Inclination" (Sep 2023) https://t.co/rjOYJVFS9H (Dec 2023) High Capacity Memory https://t.co/NrDgERzT7l High Cap Mem for Power Delivery https://t.co/mMgmSFhWHP https://t.co/we3S9U15KE J. Fryman https://t.co/DALMNfWjz5 https://x.com/ogawa_tter/status/2022156337065734249 = Vertically" [X Link](https://x.com/ogawa_tter/status/2022237491085140290) 2026-02-13T09:12Z [----] followers, [--] engagements "@1p_semicon ( Intel [-----] "Full duplex DRAM for Tightly Coupled Compute Die and Memory Die" Randy B. Osborne HASWELL eDRAM (Full duplex DRAM) ZAM ( https://x.com/ogawa_tter/status/2022175488782086654 = Systems & Methods for HBM with Unidirectional Data Flow Google (P: Aug 2023) https://t.co/vt6AOFo1Ef Concurrent R and W Intel Appls Full duplex DRAM https://t.co/7twloembBI Vertically Stacked Dies https://t.co/yfXsOZf44i Split Write DRAM MEMSYS [----] https://t.co/qrALtS0MlL https://t.co/uBOuv33PFD https://x.com/ogawa_tter/status/2022175488782086654 = Systems & Methods for HBM with" [X Link](https://x.com/ogawa_tter/status/2022237850511741235) 2026-02-13T09:14Z [----] followers, [--] engagements "@1p_semicon ( [---] Mueon Co-founder & CEO W. Gomes Intel FellowPVC/Vertically Stacked DiesIntel Capital $15.5M Seed "Cubelets" Vertically Stacked Dies https://x.com/ogawa_tter/status/2021935819871899826 = Mueon Emerges From Stealth with $15.5M Seed to Redefine DCs for the AI Era Sep [--] [----] https://t.co/CpX6V2wK48 Cubelets Wilfred Gomes Co-founder & CEO Ponte Vecchio https://t.co/7K92LswVpn Vertically Stacked Dies Intel Appls https://t.co/mMgmSFhWHP https://t.co/we3S9U15KE https://t.co/jI4bznImdt https://x.com/ogawa_tter/status/2021935819871899826 = Mueon Emerges From Stealth with $15.5M Seed" [X Link](https://x.com/ogawa_tter/status/2022238502038188344) 2026-02-13T09:16Z [----] followers, [--] engagements "= [--------] SAIMEMORY202623 NEXT25 FugakuNEXT MONAKA-X w/ Intel ZAM & Full duplex DRAM +NVIDIA GPU Stacked LLC (GC Memory) https://x.com/ogawa_tter/status/2015416350328656371 https://x.com/ogawa_tter/status/2021503395916546317 https://x.com/ogawa_tter/status/2022156337065734249 https://x.com/ogawa_tter/status/2020398753103093990 = "Three Dimensional Chip and Package Integration with Backside Metallization" NVIDIA (Sep 2024) https://t.co/IgRg0rTB4J https://t.co/ZeNTjatJiT COPA-GPU: Patent Mar [----] (P: Feb 2021) / ACM TACO Face-to-Face Dies Patents (Jan 2020) GC Memory IEDM25" [X Link](https://x.com/ogawa_tter/status/2022288800010158463) 2026-02-13T12:36Z [----] followers, [---] engagements "= Neurophos: $110M Series A to Launch Exaflop-Scale Photonic AI Chips Jan [--] [----] Nov [----] $7.2M Seed Dec [--] [----] (P: Jan 2024) Sep [----] (P: Aug 2019) https://patents.google.com/patent/US11120332B2/en https://patents.google.com/patent/WO2025151899A1/en https://otc.duke.edu/news/neurophos-raises-7-2-million-round-to-create-super-fast-energy-efficient-metamaterial-based-optical-ai-chips-for-data-centers/ https://www.neurophos.com/whitepapers https://www.neurophos.com/110m-raise https://patents.google.com/patent/US11120332B2/en https://patents.google.com/patent/WO2025151899A1/en" [X Link](https://x.com/ogawa_tter/status/2015697260898906207) 2026-01-26T08:04Z [----] followers, [---] engagements "= "Recent Advances and Trends in Hybrid Bonding" John H Lau Unimicron Technology IEEE Hybrid Bonding Symposium Jan [--] [----] (29:21) State of the Art and Outlooks of CuCu HB J. Lau. IEEE TCPMT Oct [--] https://ieeexplore.ieee.org/abstract/document/11218187 https://attend.ieee.org/hbs/wp-content/uploads/sites/692/2026-1-1-Lau.pdf https://ieeetv.ieee.org/hbs/recent-advances-and-trends-in-hybrid-bonding https://ieeexplore.ieee.org/abstract/document/11218187 https://attend.ieee.org/hbs/wp-content/uploads/sites/692/2026-1-1-Lau.pdf" [X Link](https://x.com/ogawa_tter/status/2016346912954495357) 2026-01-28T03:05Z [----] followers, [----] engagements "= "Co-Packaged Optics - Heterogeneous Integration of Chiplets (Switch Photonic IC and Electronic IC)" John H Lau Unimicron J of Microelectronics & Elect Pkg Dec [--] [----] Co-Packaged Optics Book SemiAnalysis Jan [--] [----] https://x.com/ogawa_tter/status/2007777752204898431 https://imapsjmep.org/article/154275-co-packaged-optics-heterogeneous-integration-of-chiplets-switch-photonic-ic-and-electronic-ic = Co-Packaged Optics (CPO) Book: Scaling with Light for the Next Wave of Interconnect Jan [--] [----] @SemiAnalysis_ https://t.co/TT1tvvbkiZ 1: CPO TCO Analysis 2: CPO Intro & Implementation 3: Bringing" [X Link](https://x.com/ogawa_tter/status/2017148387603472872) 2026-01-30T08:10Z [----] followers, [---] engagements "= "Xscape Photonics and Tower Semiconductor Unveil the Industrys First Optically Pumped On-Chip Multi-Wavelength Laser Platform for AI Datacenter Fabrics" Aug [--] [----] OCP OSAI Sep [--] Appl (Oct 2023) https://x.com/ogawa_tter/status/1923670710142239136 https://x.com/ogawa_tter/status/1965784807604105442 https://towersemi.com/2025/08/25/08252025/ Next-Gen DRAM [----] Mar [----] Yole Gr https://t.co/t8co4SCF4B = System-Level WDM Switching for High Bandwidth & High-Capacity Memory Access Xscape Photonics (Oct 2023) https://t.co/a99PT5455b Oct [----] https://t.co/14KFeWqwew $44M Series A Oct [--] Google" [X Link](https://x.com/ogawa_tter/status/1985723507389710618) 2025-11-04T14:59Z [----] followers, [----] engagements "= IOWN APN2026113 NTT : IOWN [--------] [-----] https://x.com/ogawa_tter/status/1886914632964514119 https://journal.ntt.co.jp/article/37949 https://journal.ntt.co.jp/article/38187 https://www.soumu.go.jp/main_content/001050373.pdf = "TSMC NTT"202523 https://t.co/UMg1egcORN NTT20241023 https://t.co/21VDLp5fnZ https://t.co/8Ocjc8PQHg IOWN1213 https://t.co/Thl7cms3Kd Dec [--] / Jul [--] https://t.co/veeVrTY0s3 https://x.com/ogawa_tter/status/1886914632964514119 https://journal.ntt.co.jp/article/37949 https://journal.ntt.co.jp/article/38187 https://www.soumu.go.jp/main_content/001050373.pdf = "TSMC" [X Link](https://x.com/ogawa_tter/status/2018252456531234831) 2026-02-02T09:17Z [----] followers, [---] engagements "= GRAPH [---] June [----] BFS 1: Supercomputer Fugaku () [----] GTEPS: [-----] (K: 31302.4) Fujitsu A64FX Tofu-D 2: Sunway TaihuLight [----] [-------] 3: Summit (CPU-Only) [----] [------] SSSP 1: SuperMUC-NG 2: NERSC Cori [----] Haswell 3; Cray CS500 Xeon Phi http://graph500.org/ http://graph500.org/" [X Link](https://x.com/ogawa_tter/status/1274989132368969729) 2020-06-22T08:54Z [----] followers, [--] engagements "= Advanced Packaging AMD Case Study Tutorial Hot Chips [----] Aug [--] (25 pp) TSMC/Intel Chiplet G. Loh Aug [--] 3D Patent EFB Nov [--] https://x.com/ogawa_tter/status/1457819506458255365 https://x.com/ogawa_tter/status/1450522719536091138 https://x.com/ogawa_tter/status/1464220199020761093 https://x.com/ogawa_tter/status/1470641381458984965 https://hc33.hotchips.org/assets/program/tutorials/2021%20Hot%20Chips%20AMD%20Advanced%20Packaging%20Swaminathan%20Final%20%2020210820.pdf https://youtu.be/csAPSw11INUt=2193 = AMD Nov [--] [----] https://t.co/tltY8wXDoq AMD Instinct MI200: Elevated Fanout Bridge 2.5D" [X Link](https://x.com/ogawa_tter/status/1470644249104470018) 2021-12-14T06:38Z [----] followers, [--] engagements ""Beyond [---] Gb/s High-speed Optical (and Copper) Interconnects" Samtec Invited TWEPP Sep [--] [----] PPTX (35 MB) https://indico.cern.ch/event/608587/contributions/2697522/attachments/1523073/2380259/TWEPP_2017_Samtec_Presentation_03.pptx https://indico.cern.ch/event/608587/contributions/2697522/attachments/1523073/2380259/TWEPP_2017_Samtec_Presentation_03.pptx" [X Link](https://x.com/anyuser/status/909060061732806656) 2017-09-16T14:23Z [----] followers, [--] engagements "= "Integrating NVIDIA Deep Learning Accelerator (NVDLA) with RISC-V SoC on FireSim" 2nd WS on Energy Efficient Machine Learning and Cognitive Computing for Embedded Applications (EMC2) Fe [--] [----] FireSim https://fires.im/ https://github.com/CSL-KU/firesim-nvdla https://docs.wixstatic.com/ugd/42e7ad_480c39fe4d4b4d7e878c8ded85041ad5.pdf https://fires.im/ https://github.com/CSL-KU/firesim-nvdla https://docs.wixstatic.com/ugd/42e7ad_480c39fe4d4b4d7e878c8ded85041ad5.pdf" [X Link](https://x.com/ogawa_tter/status/1097247036544409601) 2019-02-17T21:30Z [----] followers, [--] engagements "= "Shasta Hardware Technical Workshop" Cray CUG [----] May [--] [----] PDF [--] pp Mountain / River (19" Rack) Slingshot NIC: "Mellanox" Rosetta: Broadcom IP Exascale Day 2019/1018 Steve Scott Sep [----] = https://x.com/ogawa_tter/status/1171793483889119232 https://x.com/ogawa_tter/status/1186759552114388992 https://cug.org/proceedings/cug2019_proceedings/includes/files/inv112s1-file1.pdf = "The Cray Shasta Architecture" Steve Scott Cray HPC User Forum Sep [--] [----] https://t.co/iI95MNXz8c Slingshot: Gnodal Patents (Hot Interconnects 2019) https://t.co/yvJX3cTdmk NERSC Perlmutter https://t.co/sb3VqGAbdx" [X Link](https://x.com/ogawa_tter/status/1196034168234246144) 2019-11-17T11:55Z [----] followers, [--] engagements "= "GPUDirect Storage: Transfer Data Directly to GPU Memory Alleviating IO Bottlenecks" NVIDIA SC19 Nov [--] [----] Video Slides RAPIDS Sep [----] GPUDirect Storage Blog Aug [----] https://x.com/ogawa_tter/status/1159401097200459776 https://x.com/ogawa_tter/status/1176605978227142656 https://on-demand.gputechconf.com/supercomputing/2019/pdf/sc1922-gpudirect-storage-transfer-data-directly-to-gpu-memory-alleviating-io-bottlenecks.pdf https://on-demand.gputechconf.com/supercomputing/2019/video/sc1922-gpudirect-storage-transfer-data-directly-to-gpu-memory-alleviating-io-bottlenecks/ = "GPUDirect Storage:" [X Link](https://x.com/ogawa_tter/status/1205310718196633600) 2019-12-13T02:17Z [----] followers, [--] engagements "= ""JAMSTEC201012 PDF via @keaton_n18 Computers in Physics 1992/88 Supercomputing in Japan Oyanagi [----] [-----] https://x.com/ogawa_tter/status/1220575846206857216 https://x.com/ogawa_tter/status/1219333793674170368 https://x.com/ogawa_tter/status/1215750543731068928 http://www.jamstec.go.jp/es/jp/publication/pdf/Development_ES.pdf https://x.com/ogawa_tter/status/1220575846206857216 https://x.com/ogawa_tter/status/1219333793674170368 https://x.com/ogawa_tter/status/1215750543731068928 http://www.jamstec.go.jp/es/jp/publication/pdf/Development_ES.pdf" [X Link](https://x.com/anyuser/status/1275672840105328640) 2020-06-24T06:11Z [----] followers, [--] engagements "= "PIUMA: Programmable Integrated Unified Memory Architecture" Intel arXiv Oct [--] [----] A graph analysis oriented architecture Custom RISC ISA Hierarchical topology (HyperX) PIUMA IEEE HPEC [----] PUMA (DARPA HIVE) FOSDEM [----] https://x.com/ogawa_tter/status/1308870123973943296 https://arxiv.org/abs/2010.06277 = "Hash Table Scalability on Intel PIUMA (Programmable and Integrated Unied Memory Architecture)" Intel IEEE HPEC [----] Sep [--] [----] Efficient Sparse Matrix-Vector Multiplication on Intel PIUMA IEEE HPEC Sep [--] https://t.co/6tXhE80qbz PUMA (DARPA HIVE) Intel FOSDEM 2020" [X Link](https://x.com/ogawa_tter/status/1319097871954706434) 2020-10-22T02:06Z [----] followers, [--] engagements "= @LightOnIO Appliance a Photonic Co-processor sets a new pathway for Transformative AI / HPC computing Mar [--] [----] The First Commercial Photonic Computing in your own DC [--] TOPS/W https://lighton.ai/publications/ https://lighton.ai/wp-content/uploads/2021/03/LightOn-Appliance-spec-sheet.pdf https://lighton.ai/lighton-appliance/ https://lighton.ai/wp-content/uploads/2021/03/LightOn-Appliance-press-release.pdf https://lighton.ai/publications/ https://lighton.ai/wp-content/uploads/2021/03/LightOn-Appliance-spec-sheet.pdf https://lighton.ai/lighton-appliance/" [X Link](https://x.com/ogawa_tter/status/1368931295124811779) 2021-03-08T14:27Z [----] followers, [--] engagements "= AArch64-ExploreExploration of Apple CPUs @handleym99 Vol 1: CPU Core (205 pp) Vol 2: Load Store Unit Cache & Memory (215 pp) Vol 3: Patent Exploration (116 pp) My Apple patent tweets https://twilog.org/ogawa_tter/searchword=Apple%20patent&ao=a https://github.com/name99-org/AArch64-Explore https://x.com/handleym99/status/1511475655447232512 OK folks the big update is here Everything can be found at https://t.co/HDozWzQRHX I know nothing about Git or Github and have little interest in learning so I likely screwed up something but what you should find here is 1/" [X Link](https://x.com/ogawa_tter/status/1511558036032147456) 2022-04-06T04:14Z [----] followers, [--] engagements "= Intel & SoftBank Subsidiary SAIMEMORY Collaborate to Advance Next-Generation Memory for AI Feb [--] [----] Z-Angle Memory Jan [--] AMT Dec [----] Appl (Dec 2023) https://x.com/ogawa_tter/status/2019040243077570752 https://x.com/ogawa_tter/status/1603609370859737089 https://www.sandia.gov/research/news/from-early-rd-to-mature-technology/ https://eetimes.itmedia.co.jp/ee/articles/2602/03/news099.html https://community.intel.com/t5/Blogs/Intel/Policy-Intel/Intel-and-SoftBank-Subsidiary-SAIMEMORY-Collaborate-to-Advance/post/1735559 = "Package Architectures having Vertically Stacked Dies for High Capacity" [X Link](https://x.com/ogawa_tter/status/2019041233503658372) 2026-02-04T13:31Z [----] followers, [----] engagements ""SAIMEMORY"202623 [----] Eng = Intel Feb [--] EE Times [--] AMT (NNSA/Sandia) Jan [--] Intel Appl (Dec 2023) https://x.com/ogawa_tter/status/2019041233503658372 https://www.saimemory.co.jp/en/news/press/20260203_01/ https://www.saimemory.co.jp/news/press/20260203_01/ = Intel & SoftBank Subsidiary SAIMEMORY Collaborate to Advance Next-Generation Memory for AI Feb [--] [----] https://t.co/uIHQPTh4bb https://t.co/EFa0mj0mL7 Z-Angle Memory Jan [--] https://t.co/4HF7hLvQcm AMT Dec [----] https://t.co/ch6fI6A3Vk Appl (Dec 2023) https://t.co/NrDgERzT7l https://t.co/j3U50UkvW0" [X Link](https://x.com/ogawa_tter/status/2019249282877125075) 2026-02-05T03:18Z [----] followers, [----] engagements "= Vertically Stacked Dies Intel Appls Stephen Morein CTO SAIMEMORY (Aug [----] --) w/ Solder Interconnects (Aug 2023) (Jun 2022) C Bricks Planarized Edges = https://patents.google.com/patent/US20230420409A1/en https://patents.google.com/patent/US20230420411A1/en https://patents.google.com/patent/US20240006375A1/en https://patents.google.com/patent/US20250062278A1/en https://x.com/ogawa_tter/status/2022156337065734249 = Vertically Stacked Dies Intel Appls "Memory Layers Bonded to Logic Layers with Inclination" (Sep 2023) https://t.co/rjOYJVFS9H (Dec 2023) High Capacity Memory" [X Link](https://x.com/ogawa_tter/status/2022579290274144512) 2026-02-14T07:50Z [----] followers, [---] engagements "= "Double Coil Inductive Scheme and Modulation for Z-Axis Memory (ZAM) Links" Intel Appl (Jul 2024) Stephen Morein CTO SAIMEMORY (Aug [----] --) SAIMEMORY Feb [--] [----] ZAM (Z-Angle Memory) AMT NNSA Jan [--] https://x.com/ogawa_tter/status/2019249282877125075 https://x.com/ogawa_tter/status/2022579290274144512 https://patents.google.com/patent/US20260005164A1/en "SAIMEMORY"202623 https://t.co/O2FnHYEy8C [----] Eng https://t.co/YgWteRQQZV = https://t.co/B6ghH06aS2 Intel Feb [--] EE Times [--] AMT (NNSA/Sandia) Jan [--] Intel Appl (Dec 2023) https://t.co/hhekUv4KS8" [X Link](https://x.com/ogawa_tter/status/2022579523662025054) 2026-02-14T07:51Z [----] followers, [---] engagements "Brain-Inspired Intelligent SoCs and Applications Hoi-Jun Yoo KAIST Keynote MCSoC-16 PDF http://ssl.kaist.ac.kr/2007/ http://mcsoc-forum.org/2016/wp-content/uploads/2015/07/keynote-Yoo.pdf http://ssl.kaist.ac.kr/2007/ http://mcsoc-forum.org/2016/wp-content/uploads/2015/07/keynote-Yoo.pdf" [X Link](https://x.com/anyuser/status/826632295083712512) 2017-02-01T03:24Z [----] followers, [--] engagements "= "Stream-Dataflow Acceleration" ISCA [----] Slides [---] pages Paper RISC-V based simulator https://x.com/ogawa_tter/status/883310226958450688 http://research.cs.wisc.edu/vertical/talks/2017/isca17-stream-dataflow-talk = "Stream-Dataflow Acceleration" Tony Nowatzki et al. ISCA [----] PDF https://t.co/LP3gHqOru3 Softbrain Cycle-level RISC-V based simulator https://t.co/H7LCvR923y https://x.com/ogawa_tter/status/883310226958450688 http://research.cs.wisc.edu/vertical/talks/2017/isca17-stream-dataflow-talk = "Stream-Dataflow Acceleration" Tony Nowatzki et al. ISCA [----] PDF https://t.co/LP3gHqOru3" [X Link](https://x.com/ogawa_tter/status/894887721251987456) 2017-08-08T11:47Z [----] followers, [--] engagements "= "A Deep Dive Into NEC's Aurora Vector Engine" Nov [--] [----] MPI operations to be directory executed between the VEs w/o copying VE can access Xeon DDR4 via DMA 6x HBM2 1.2TB/s VecLen: [---] x 64bit 2.45TF @ 1.6GHz STREAM/node price HPL/node price https://www.nextplatform.com/2017/11/22/deep-dive-necs-aurora-vector-engine/ https://www.nextplatform.com/2017/11/22/deep-dive-necs-aurora-vector-engine/" [X Link](https://x.com/ogawa_tter/status/933484240154398720) 2017-11-22T23:56Z [----] followers, [--] engagements "= "A case for scoped persist barriers in GPU" AMD Research GPGPU [----] Feb [--] [----] HW support for persist barriers The only additional HW is Non-Volatile-Write FIFOs AMD Radeon Pro SSG Datasheet HBCC http://pro.radeon.com/_downloads/radeon-pro-ssg-datasheet-27.7.17.pdf https://pro.radeon.com/en/product/pro-series/radeon-pro-ssg/ http://drona.csa.iisc.ac.in/arkapravab/papers/gpgpu18_scoped_persistent_barriers.pdf http://pro.radeon.com/_downloads/radeon-pro-ssg-datasheet-27.7.17.pdf https://pro.radeon.com/en/product/pro-series/radeon-pro-ssg/" [X Link](https://x.com/ogawa_tter/status/970984947321208833) 2018-03-06T11:30Z [----] followers, [--] engagements ""Tesla's Autopilot Chief Keller Steps Down After Two Years" Bloomberg Apr [--] [----] 12:17 JST Updated 13:18 JST "AMD CPU hero Jim Keller leaving Tesla to join Intel" TweakTown Apr [--] [----] JST "According to my industry sources" https://www.tweaktown.com/news/61675/amd-cpu-hero-jim-keller-leaving-tesla-join-intel/index.html https://www.bloomberg.com/news/articles/2018-04-26/tesla-says-autopilot-vice-president-keller-is-leaving-carmaker https://www.tweaktown.com/news/61675/amd-cpu-hero-jim-keller-leaving-tesla-join-intel/index.html" [X Link](https://x.com/ogawa_tter/status/989384713952346114) 2018-04-26T06:04Z [----] followers, [--] engagements "= Qualcomm Acquires Ventana Micro Systems Deepening #RISCV CPU Expertise Dec [--] [----] "Ventana team will complement Qualcomms ongoing RISC-V and custom Oryon CPU development further advancing technology leadership in the AI era across all businesses." https://www.qualcomm.com/news/releases/2025/12/qualcomm-acquires-ventana-micro-systems--deepening-risc-v-cpu-ex https://www.qualcomm.com/news/releases/2025/12/qualcomm-acquires-ventana-micro-systems--deepening-risc-v-cpu-ex" [X Link](https://x.com/ogawa_tter/status/1998940542835814660) 2025-12-11T02:18Z [----] followers, [---] engagements "= ""NRIJST2025822 20258JICA (314) HPC https://x.com/ogawa_tter/status/1993869856249418069 https://x.com/ogawa_tter/status/1986425912309784825 https://spap.jst.go.jp/event/pdf/study045.pdf https://www.youtube.com/watchv=2lHaOqMcnnY = "India has satisfied its supercomputing needs but not its ambitions" Nov [--] [----] https://t.co/ALa8rlROEG [--] machines: [--] PF + [--] PF [----] National Supercomputing Mission RSDs [----] https://t.co/3zQC5ZMvtk Rudra Server AUM 5nm [--] cores: 2x 24x (2x Neoverse-V2) DDR5 HBM3 https://x.com/ogawa_tter/status/1993869856249418069" [X Link](https://x.com/ogawa_tter/status/1998986106570944734) 2025-12-11T05:20Z [----] followers, [---] engagements "= Memory: Almost The Only Thing That Matters A Revolution in Memory Arch for the DC Marvell Hot Chips [----] Custom SRAM Jun [--] Custom HBM Dec [----] [--] Gbps/wire BiDI D2D Aug [--] https://www.marvell.com/company/newsroom/marvell-unveils-64gbps-bidirectional-die-to-die-interface-ip-2nm-xpu.html https://x.com/ogawa_tter/status/1935576279216197659 https://hc2025.hotchips.org/assets/program/conference/day2/40_Marvell_Kuemerle_final.pdf https://www.youtube.com/watchv=p97UkNh2oPc&t=38s = Marvell Custom AI Investor Event Jun [--] [----] (2:32:41) https://t.co/F2VESxFs4t (124 pp) https://t.co/qPrcAtqXP0" [X Link](https://x.com/ogawa_tter/status/1999123077238751525) 2025-12-11T14:24Z [----] followers, [---] engagements "= "Accelerating AI with UALink: Open Memory Fabrics for Scalable Compute" R. Hyatt UnifabriX Dec [--] [----] Alibaba: ALink AMD Helios & HPE: UALoE UnifabriX https://x.com/ogawa_tter/status/1967956344792813837 https://x.com/ogawa_tter/status/1996037944759746638 https://x.com/ogawa_tter/status/1986634590325514505 https://ualinkconsortium.org/wp-content/uploads/2025/12/UnifabriX_UALink-Webminar_20251210_PDF.pdf https://www.youtube.com/watchv=N_8eVUd0utA = UnifabriX https://t.co/k5BiHYCMw0 Out of Stealth Mode Nov [--] [----] https://t.co/PP26HbEWrJ Lord of the Memory Fabrics ISW [----] Sep 10" [X Link](https://x.com/ogawa_tter/status/1999354746826858848) 2025-12-12T05:44Z [----] followers, [---] engagements "= United States and Japan Will Sign Pax Silica Preamble Dec [--] [----] Pax Silica Initiative: Fact Sheet Pax Silica Summit & Sony Samsung SK Hynix ASML etc https://www.state.gov/pax-silica/ https://www.state.gov/releases/office-of-the-spokesperson/2025/12/pax-silica-initiative/ https://x.com/UnderSecE/status/1999315601290822037 https://www.state.gov/releases/office-of-the-spokesperson/2025/12/united-states-and-japan-will-sign-pax-silica-preamble/ Honored to spearhead todays landmark summit in collaboration with @DeputySecState @DavidSacks47 Ethan Klein alongside other colleagues across the" [X Link](https://x.com/ogawa_tter/status/1999435401799565450) 2025-12-12T11:05Z [----] followers, [---] engagements "= "NeuronMM: High-Performance Matrix Multiplication for LLM Inference on AWS Trainium" arXiv Nov [--] (Oct 29) PSUM w/ near-memory accumulation to enables read-accumulate-write ops Trainium3 NeuronCore-v4 https://awsdocs-neuron.readthedocs-hosted.com/en/latest/about-neuron/arch/neuron-hardware/neuron-core-v4.html https://x.com/ogawa_tter/status/1997870805313876448 https://arxiv.org/abs/2510.25977 = AWS Trn3 UltraServers: Power next-gen enterprise AI performance AWS re:Invent [----] https://t.co/sqC5x5Ty0a HWed Microslcaling (MXFP8) & Softmax func Trainium4 NVIDIA Dec [--] https://t.co/k3M2NMSfn9" [X Link](https://x.com/ogawa_tter/status/1999680466727940181) 2025-12-13T03:19Z [----] followers, [---] engagements "= "Made in China 2025: Evaluating China's Performance" Nov [--] [----] U.S.-China Economic & Security Review Commission Scorecard Next-Gen Information Technology: Semiconductors . Made in China [----] [----] May [----] https://x.com/ogawa_tter/status/1501989064651640842 https://www.uscc.gov/research/made-china-2025-evaluating-chinas-performance = Translation: Notice of the State Council on the Publication of Made in China [----] CSET Mar [--] [----] https://t.co/PGfrToSPNh One of Chinas most important industrial policies [----] [------] https://t.co/wPwuZsBxFF https://t.co/u32cUVdwX2 https://t.co/yldNiZN6sL" [X Link](https://x.com/ogawa_tter/status/1999746724773134495) 2025-12-13T07:42Z [----] followers, [---] engagements "= "Magnetics Integrations for 2.5D and 3D Packaging" Tyndall Power Magnetics @ High Frequency [----] WS Mar [--] The necessity of 2.5D/3D integration of magnetics Classification of 3D integrated magnetics 2.5D HI of magnetics by micro-transfer technology https://psma.com/sites/default/files/uploads/MagneticsWorkshopPresentations/Magnetics_Integrations_for_2.5D_and_3D_Packaging__Sai_Tyndall_S.pdf https://psma.com/sites/default/files/uploads/MagneticsWorkshopPresentations/Magnetics_Integrations_for_2.5D_and_3D_Packaging__Sai_Tyndall_S.pdf" [X Link](https://x.com/ogawa_tter/status/1999815958077354480) 2025-12-13T12:17Z [----] followers, [---] engagements "= "Reimaging and Transforming Package Assembly and Test Manufacturing in the AI Era" SATAS () VP Foundry Tech Manufacturing Intel HIR Session at SEMICON West [----] Oct [--] NEDO [------] https://x.com/ogawa_tter/status/1854410029757329749 https://eps.ieee.org/images/files/HIR_Presentations/SEMICON_W_2025/Oct_8_Jeffrey_Pattinato.pdf https://x.com/ogawa_tter/status/1854410029757329749 https://eps.ieee.org/images/files/HIR_Presentations/SEMICON_W_2025/Oct_8_Jeffrey_Pattinato.pdf" [X Link](https://x.com/ogawa_tter/status/2000053660810981472) 2025-12-14T04:02Z [----] followers, [--] engagements "= "Feeding the Beast: The Critical Challenges of Powering Next-Gen AI accelerators" Jawad Haj-Yahya Rivos 1st @SAFARI_ETH_CMU Conf [----] Dec [--] @_onurmutlu_ DC [---] VDC Vertical PD https://x.com/ogawa_tter/status/1973657582549213514 https://x.com/ogawa_tter/status/1981369625276551328 https://events.safari.ethz.ch/SafariConference25/lib/exe/fetch.phpmedia=jawad_future_of_pdn_2025.pptx https://www.youtube.com/live/QtS7v8aVGLQt=11530s = "Hybrid Switched-Capacitor Circuits and Magnetics Co-Design for Vertical Power Delivery" PhD Thesis [----] https://t.co/sHbynIDNNU Mini-LEGO Nov 2023" [X Link](https://x.com/ogawa_tter/status/2000906835470209427) 2025-12-16T12:32Z [----] followers, [---] engagements "= Federal gov terminates $285M contract for SMART USA CHIPS manufacturing institute Dec [--] [----] 'termination for convenience' Executive Director ORAU Webinar Oct [--] Feb [--] https://x.com/ogawa_tter/status/1963953570593771548 https://www.orau.org/partnerships/documents/webinar/digital-manufacturing-smart.pdf https://www.smartusatwins.org/ https://www.wral.com/news/local/federal-terminates-285m-contract-durham-chips-institute-december-2025/ = Digital Twin in Semicon How AI/ML Advances HI Keynote HIR Sessions ECTC [----] https://t.co/TAVObdxKbY SMART USA HIR Feb [--] https://t.co/gj8L6UeEHr Jan 3" [X Link](https://x.com/ogawa_tter/status/2001156497288613985) 2025-12-17T05:04Z [----] followers, [----] engagements "= "Chip Scan: 3D X-ray imaging of CMOS integrated circuits" A.F.J. Levi USC Keynote Microelectronics Reliability and Qualification WS (MRQW) Feb [--] [----] 4nm Nature Jul [--] [----] 22nm Apr [--] [----] https://x.com/ogawa_tter/status/1818874067765936260 https://www.nature.com/articles/s41586-024-07615-6.epdfsharing_token=wRppPTjk7NeAyZocQG2NkdRgN0jAjWel9jnR3ZoTv0MFXlcEC43QR4AJLMRV3QQH27tVRSk9jZqg2AuLkhsTaedXagMSnk7I3F3m2gJfmEJMcrxLcjnmlxxcZB5kOgr51a6rPpBo-UJ3x2dBdbogakoNROieAnvF1oQj1wDCpsk%3D" [X Link](https://x.com/ogawa_tter/status/2001285742425972760) 2025-12-17T13:37Z [----] followers, [---] engagements "= "Solving The Problems of HBM-on-Logic" Dec [--] [----] @gavbon86 (Guest) & @IanCutress imec at IEDM [----] Future AI Accelerators might Need to be Slower to be Faster HIFOM (Hynix Integrated Fan-out Memory) Logic on stacked DRAM https://x.com/ogawa_tter/status/1726944199168430108 https://morethanmoore.substack.com/p/solving-the-problems-of-hbm-on-logic = "Fan-out Wafer Level Package for Memory Applications" SK hynix Chip Scale Review Jan-Feb [----] https://t.co/norLW2GJYu ECTC [----] https://t.co/MpzVX2ZUD9 Patent May [----] (KR: Aug 2018) https://t.co/L5dfZew0vQ SK hynix Patent Apr [----] (KR: Aug 2020)" [X Link](https://x.com/ogawa_tter/status/2001503290513891468) 2025-12-18T04:02Z [----] followers, [---] engagements "= Loongson performance in benchmarks Tramplin Electronics and NEIMARK: LoongArch Architecture Meetup Dec [--] [----] Tramplin Electronics unveiled the Russian Irtysh C616 server processor Nov [--] based on LoongArch https://x.com/ogawa_tter/status/1995737047512793284 https://vk.com/neimark_itw=wall-211802468_1420 = Tramplin Electronics unveiled the Russian Irtysh C616 server processor Nov [--] [----] https://t.co/E5u7wxHW5G based on LoongArch architecture C616: [--] cores [---] GHz [-----] GF 632: . 664: [--] cores [---] GHz [----] GF https://t.co/wXQY20oIDx https://t.co/91L8nl7qsX https://t.co/oAjg6pHY0L" [X Link](https://x.com/ogawa_tter/status/2001557252290023688) 2025-12-18T07:36Z [----] followers, [---] engagements "= SCREEN Hokdings IR Day for FY2026/03 Dec [--] [----] ATCA: $75M Strategic Partnership w/ NY Creates IBM Sep [--] Nikon Oct [--] [-----] IR Day1216 https://www.screen.co.jp/download_file/view/88a2eb9f-6bcc-4ff6-acfb-7bea28f88252/449 https://www.screen.co.jp/en/news/NR251031E https://www.screen.co.jp/spe/en/information/spe250924 https://ny-creates.org/governor-hochul-announces-75-million-strategic-partnership-between-ny-creates-and-screen-to-strengthen-u-s-and-japan-semiconductor-rd-collaboration/ https://www.screen.co.jp/download_file/view/c840ea6a-82e5-4a4a-a930-77e0a8dc8d3f/1210" [X Link](https://x.com/ogawa_tter/status/2001635246962512109) 2025-12-18T12:46Z [----] followers, [----] engagements "= "Semiconductor challenges in ramping AI at the edge" PR Chidi Chidambaram SVP Engineering & Qualcomm Fellow Qualcomm Technologies HIR Session at SEMICON West [----] Oct [--] Device 3D + System 3D Hetero. integration STCO: New Language Needed https://eps.ieee.org/images/files/HIR_Presentations/SEMICON_W_2025/Oct_8_PR_Chidi_Chidambaram.pdf https://eps.ieee.org/images/files/HIR_Presentations/SEMICON_W_2025/Oct_8_PR_Chidi_Chidambaram.pdf" [X Link](https://x.com/ogawa_tter/status/2001858926548258976) 2025-12-19T03:35Z [----] followers, [---] engagements "= "Survey of storage systems in high performance computing" NUDT China Computer Federation (CCF) Transactions on HPC Dec [--] [----] Systematically reviews covering four aspects: Storage architecture Hardware Software and Networking [--] pp [--] () ref https://link.springer.com/article/10.1007/s42514-025-00268-5 https://link.springer.com/article/10.1007/s42514-025-00268-5" [X Link](https://x.com/ogawa_tter/status/2001915837108953315) 2025-12-19T07:21Z [----] followers, [---] engagements ""Photonic Connectivity for Al Systems" K. Bergman. Columbia EFCL Int WS on Sustainable AI Nov [--] [----] 3D Photo OXC Jul [--] Beyond CLOS Huawei = HBM Columbia HotI [----] https://x.com/ogawa_tter/status/2001896001653932197 https://x.com/ogawa_tter/status/1985550170873122966 https://x.com/ogawa_tter/status/1904371742581408209 https://video.ethz.ch/events/2025/efcl/v/Jp2fo7TbBhw = "Solving The Problems of HBM-on-Logic" Dec [--] [----] https://t.co/QET8P02fUb HBM (Memory) with an optical connection Celestial AI https://t.co/FhgK8qU2Jb IEEE CAL https://t.co/HZ5uxuLmx3 Google Appls Columbia HotI 2025" [X Link](https://x.com/ogawa_tter/status/2002006934044680479) 2025-12-19T13:23Z [----] followers, [---] engagements "= ACTINA: Adapting Circuit-Switching Techniques for AI Networking Arch Columbia & NVIDIA SC25 Flexible Photo Accelerated High Performance Compute System Z. Wu PhD Thesis [----] = NVIDIA K. Bergman Nov [--] https://x.com/ogawa_tter/status/2002006934044680479 https://academiccommons.columbia.edu/doi/10.7916/m381-v267 https://dl.acm.org/doi/full/10.1145/3712285.3759842 "Photonic Connectivity for Al Systems" K. Bergman. Columbia EFCL Int WS on Sustainable AI Nov [--] [----] https://t.co/l9ifnh9xfg 3D Photo https://t.co/Zc7KbYUv9g OXC Jul [--] https://t.co/mJ6CjRoi7r Beyond CLOS Huawei = HBM" [X Link](https://x.com/ogawa_tter/status/2002297785232863709) 2025-12-20T08:39Z [----] followers, [---] engagements "= "Modularity at the Package Edge: Composable Interconnect for [---] Gb/s/Lane AI Systems" Bijan Nowroozi Head of Ecosystem Development Lightmatter OCP HPC Sub-Proj Dec [--] [----] Next-Gen OAM 400Gbs/lane fabric at the system level https://drive.google.com/file/d/1eQbxpCDgtlM1lH9JCyxUQ2cLPFc0b_fj/view https://www.youtube.com/watchv=CaIyMse50Jo https://drive.google.com/file/d/1eQbxpCDgtlM1lH9JCyxUQ2cLPFc0b_fj/view https://www.youtube.com/watchv=CaIyMse50Jo" [X Link](https://x.com/ogawa_tter/status/2002690817467863143) 2025-12-21T10:41Z [----] followers, [---] engagements "= NVIDIA CUDA Tile using cuTile and CUDA Tile IR to write portable high-performance GPU Kernels @JokerEph & @roeschinc GPU Mode Dec [--] [----] cuTile Python Dec [--] May [--] GTC [----] https://x.com/ogawa_tter/status/1925023993679290758 https://developer.nvidia.com/blog/simplify-gpu-programming-with-nvidia-cuda-tile-in-python https://developer.nvidia.com/cuda/tile https://www.youtube.com/watchv=_b4I4rKpsGA = "The Essence of CUDA C++ : Past Present and Future" V. Grover Sr. Distinguished Engineer NVIDIA Keynote AI in the Cloud IAP May [--] [----] https://t.co/cFw7Ca6f8h Mar [----] https://t.co/TcartJ9JCu CUDA" [X Link](https://x.com/ogawa_tter/status/2002971559053639796) 2025-12-22T05:16Z [----] followers, [---] engagements "= "Split Write DRAM: Reducing DRAM Access Latency by Mitigating Read-Write Interference" K Chitra (IIT Guwahati) Minesh Patel (Rutgers) MEMSYS [----] enable timely write draining while minimizing read disruption M. Patel Dec [--] https://x.com/ogawa_tter/status/2002744203726700723 https://www.memsys.io/wp-content/uploads/ninja-forms/5/Split_Write_DRAM.pdf = "Addressing Memory Contention in Modern DRAM-Based Systems" M. Patel Rutgers SAFARI Conf Dec [--] [----] https://t.co/Al5ItSQcAN https://t.co/bXiQx1A8Q4 PhD Thesis [----] https://t.co/BiauTLdXfy @_onurmutlu_ R/W in parallel POWER9=Power10 Google" [X Link](https://x.com/ogawa_tter/status/2002994910459576421) 2025-12-22T06:49Z [----] followers, [---] engagements "= "CXLAimPod: CXL Memory is all you need in AI era" arXiv Aug [--] [----] Benefits of CXL's Full-duplex Yiwei Yang UCSC Split Write DRAM (Half) Full-duplex POWER9 Power10: Open Memory Interface Google Appl https://x.com/ogawa_tter/status/2002744203726700723 https://scholar.google.com/citationsuser=sDUT22MAAAAJ&view_op=list_works&sortby=pubdate https://arxiv.org/abs/2508.15980 = "Addressing Memory Contention in Modern DRAM-Based Systems" M. Patel Rutgers SAFARI Conf Dec [--] [----] https://t.co/Al5ItSQcAN https://t.co/bXiQx1A8Q4 PhD Thesis [----] https://t.co/BiauTLdXfy @_onurmutlu_ R/W in parallel" [X Link](https://x.com/ogawa_tter/status/2003433315152576960) 2025-12-23T11:51Z [----] followers, [---] engagements "= "Memory Access Scheduling" ISCA@50 25-Year Retrospective Jun [----] ISCA [----] S. Rixner PhD Thesis [----] William J. Dally Stream Processors Mar [----] SPI https://en.wikipedia.org/wiki/Stream_Processors_Inc https://x.com/ogawa_tter/status/1987143958276059280 https://dspace.mit.edu/handle/1721.1/86579 https://dl.acm.org/doi/10.1145/339647.339668 https://sites.coecis.cornell.edu/isca50retrospective/files/2023/06/Rixner_2000_Memory.pdf = William James Dally [----] Benjamin Franklin Medal in Computer & Cognitive Science https://t.co/OtFWa4CqZP https://t.co/WucQGgalko Merrimac SC03 Brook for GPUs" [X Link](https://x.com/ogawa_tter/status/2003473794288255328) 2025-12-23T14:32Z [----] followers, [---] engagements "= "XEM: Tensor accelerator for AB21 supercomputing artificial intelligence processor" ETRI Journal Oct [--] [----] AB21: TSMC 12-nm. [-----] x [-----] m [-----] W XEM 4x4@FP64 / 8x8@FP32 4.58TF / 18.32TF [----] MHz AB21 MUG [----] https://x.com/ogawa_tter/status/1963190097635061895 https://onlinelibrary.wiley.com/doi/10.4218/etrij.2024-0141 = Supreme-K: A Korean Initiative for Pioneering the AB21 HPC Accelerator & Compute Node ETRI MUG [----] Aug [--] https://t.co/jlFyGtrOKL https://t.co/PK3OV9alrw AB21: 8TF@FP64/32TF@FP32 TSMC 12nm #RISCV @AmpereComputing MUG [----] https://t.co/ixKwdLWaV5 K-AB21 ARM HPC UG" [X Link](https://x.com/ogawa_tter/status/2003708598527099096) 2025-12-24T06:05Z [----] followers, [---] engagements "= DoD RAMP-C Phase [--] Apr [--] [----] Prototype on Intel 18A in [----] Jul [--] [----] RAMP-C customers NVIDIA QCOM Microsoft . Intel NDIA Aug [--] [----] Reuters Dec [--] [----] 🤔🤔🤔 https://www.reuters.com/world/us/how-silicon-valley-dealmaker-charmed-trump-gave-intel-lifeline-2025-12-24/ https://x.com/IanCutress/status/2003969925602271236 https://ndia.dtic.mil/wp-content/uploads/2023/eti/Brinker.pdf https://x.com/ogawa_tter/status/1782618702846058821 Re NVIDIA no longer making test chips on 18A. IT WAS A GOVERNMENT PROGRAM to allow potential partners to build designs using PDKs. It was never a committment" [X Link](https://x.com/ogawa_tter/status/2004005284600336836) 2025-12-25T01:44Z [----] followers, [---] engagements "= Groq and NVIDIA Enter Non-Exclusive Inference Technology Licensing Agreement Dec [--] [----] J. Ross S. Madra & other members of the Groq team will join NVIDIA Groq will continue w/ S. Edwards Deterministic Memory (HBM) D. Abts https://x.com/ogawa_tter/status/1968963891481780677 https://groq.com/newsroom/groq-and-nvidia-enter-non-exclusive-inference-technology-licensing-agreement-to-accelerate-ai-inference-at-global-scale = Groq Raises $750M as Inference Demand Surges Sep [--] [----] https://t.co/VocPTQTweu "Groq next gen chip on Samsung SF4X is going to use 3D stacked SRAM" Oct 2024" [X Link](https://x.com/ogawa_tter/status/2004015056871420165) 2025-12-25T02:23Z [----] followers, [----] engagements "= Optical cross-connect (MEMS OXC) Huawei (CN: May 2024) (CN: Mar 2024) Micromirror Chips (CN: Feb 2024) (CN: Jan 2024) OXC Jul [--] Huawei OXC https://x.com/ogawa_tter/status/1985550170873122966 https://patents.google.com/patent/WO2025152651A1/en https://patents.google.com/patent/WO2025161698A1/en https://patents.google.com/patent/WO2025200814A1/en https://patents.google.com/patent/WO2025246528A1/en = Research Progress of High-Degree OXC Technologies Invited Jul [--] [----] https://t.co/qfLXQNnIK4 Huawei OptiXtrans DC808 Sep [--] [----] https://t.co/ueLHFQvKfa Special Prize at Interop Tokyo [----] Jun" [X Link](https://x.com/ogawa_tter/status/2004470061063651518) 2025-12-26T08:31Z [----] followers, [---] engagements "= Groq & NVIDIA Enter Non-Exclusive Inference Tech Licensing Agreement Dec [--] [----] J. Ross S. Madra & other members of the Groq team will join NVIDIA Dense packaging: HBM Stack Deterministic HBM D. Abts https://research.nvidia.com/person/dennis-abts https://x.com/ogawa_tter/status/1968963891481780677 https://x.com/ogawa_tter/status/2004015056871420165 = Groq Raises $750M as Inference Demand Surges Sep [--] [----] https://t.co/VocPTQTweu "Groq next gen chip on Samsung SF4X is going to use 3D stacked SRAM" Oct [----] https://t.co/zhxAJFh9wX Die-to-Die Dense Packaging (P: Feb 2022)" [X Link](https://x.com/ogawa_tter/status/2004677049685008484) 2025-12-26T22:13Z [----] followers, [---] engagements "= "Semiconductor Packaging Structure and Electronic Device" Huawei Appl (CN: Apr 2024) Stacked DRAM (400) without TSV Appls DRAM (CN: Nov 2021) Stacked DRAM (CN: Dec 2021) Proprietary HBM: HiBL / HiZQ HUAWEI CONNECT [----] Sep [--] https://x.com/ogawa_tter/status/1970494560678879673 https://patents.google.com/patent/WO2025222874A1/en = Huawei Appls DRAM (CN: Nov 2021) https://t.co/GJTecgmYes to reduce an area & power consumption Stacked DRAM (CN: Dec 2021) https://t.co/tz8dSLnjbu [--] Storage Layers Good DRAM Die: 90% Yield: 78.57% (65.61%) HUAWEI CONNECT [----] Sep [--] https://t.co/7yZaSuJsPB" [X Link](https://x.com/ogawa_tter/status/2004922428020183375) 2025-12-27T14:28Z [----] followers, [----] engagements "= "LLM Inference Scheduling: A Survey of Techniques Frameworks and Trade-offs" Huawei Canada/China Simon Fraser U. Tsinghua U. U. of British Columbia TechRxiv Nov [--] [----] Two core architectural layers: Control Plane Data Plane [--] pp [---] ref https://www.techrxiv.org/users/994660/articles/1355915-llm-inference-scheduling-a-survey-of-techniques-frameworks-and-trade-offs https://www.techrxiv.org/users/994660/articles/1355915-llm-inference-scheduling-a-survey-of-techniques-frameworks-and-trade-offs" [X Link](https://x.com/ogawa_tter/status/2005518637529276538) 2025-12-29T05:57Z [----] followers, [---] engagements "= Jul [--] [----] "An Outlook for New Homogeneous Computing System" Huawei Fellow Oct [----] Unified Bus APNet [----] Sep [--] [----] UB-Mesh HC25 SIMT / SIMD https://x.com/ogawa_tter/status/1997191844665630983 https://x.com/ogawa_tter/status/1996559876022247671 https://x.com/ogawa_tter/status/1969593294117355902 https://x.com/ogawa_tter/status/1912038192943170021 https://mp.weixin.qq.com/s/iqRXwdHjyMI7egMrXEIfFw = Graphics Processing Device and Mode Switching Method thereof Huawei Appl (CN: Feb 2023) https://t.co/NWolxy4UcQ Switching between SIMT and SIMD "Matrix Multiplier" US: May [----] (CN: Dec 2017)" [X Link](https://x.com/ogawa_tter/status/2005532420033155255) 2025-12-29T06:52Z [----] followers, [---] engagements "= Huawei Appls 3D SRAM (CN: Sep 2023) 3D 2T1C FeRAM Buffer (CN: Sep 2022) 3D 2T0C/2T2C FeRAM (CN: Aug 2022) FeRAM (CN May/Apr 2022) Stacked DRAM (CN Apr 2024) https://x.com/ogawa_tter/status/2004922428020183375 https://x.com/ogawa_tter/status/1929844531157258464 https://patents.google.com/patent/US20250185254A1/en https://patents.google.com/patent/CN117750776A/en https://patents.google.com/patent/WO2025055404A1/en = "Semiconductor Packaging Structure and Electronic Device" Huawei Appl (CN: Apr 2024) https://t.co/iZuCutuRbj Stacked DRAM (400) without TSV Appls https://t.co/PrbrXhve4d DRAM (CN:" [X Link](https://x.com/ogawa_tter/status/2005932038839574981) 2025-12-30T09:20Z [----] followers, [---] engagements "= "NVIDIA spends $5B on Intel bailout instantly gets $2.5B richer" Dec [--] [----] closed Monday at $36.68 Intel FROM-8-K Dec [--] (26) Sep [--] NVIDIA to invest $5 billion in Intel common stock $23.28/share https://x.com/ogawa_tter/status/1968660914803978356 https://www.intc.com/filings-reports/all-sec-filings/content/0000050863-25-000204/0000050863-25-000204.pdf https://www.theregister.com/2025/12/29/nvidia_intel_5_billion/ = "Intel and NVIDIA to Jointly Develop AI Infrastructure and Personal Computing Products" Sep [--] [----] https://t.co/GZuOacjYEj NVIDIA to invest $5B in Intel common stock will" [X Link](https://x.com/ogawa_tter/status/2005998170191585496) 2025-12-30T13:43Z [----] followers, [---] engagements "= Enabling Dataflow Execution on GPUs w/ Spatial Pipelines NVIDIA ACM TACO Dec [--] [----] K. Sankaralingam Sep 21- Principal Research Scientist NVIDIA Aug 17- Prof UW Madison SimpleMachines (-May 21) https://x.com/ogawa_tter/status/1602458264226045952 https://karu.sites.cs.wisc.edu/wiki/index.php https://dl.acm.org/doi/10.1145/3777466 = SimpleMachines Inc. https://t.co/djds2fJYXe What's . Karthikeyan Sankaralingam https://t.co/xmryTor5vw "founded a chip startup SimpleMachines." K. Sankaralingam et al. ISCA [----] https://t.co/8QWGUpi6x9 Oct [--] [----] https://t.co/Nj60b306AQ https://t.co/s0tYjl99em" [X Link](https://x.com/ogawa_tter/status/2006346714761117839) 2025-12-31T12:48Z [----] followers, [---] engagements ""Hardware Acceleration for Neural Networks: A Comprehensive Survey" ASU arXiv Dec [--] [----] [--] pp [--] ref S. Gupta = Zhonghao Xinying () 2nd-gen (7nm) in [----] Dec [--] : VLIW AI Chip (TPU) https://x.com/ogawa_tter/status/2005984006467379568 https://scholar.google.com/citationsuser=U9bcQkMAAAAJ&view_op=list_works&sortby=pubdate https://arxiv.org/abs/2512.23914 = Dialogue with Tech Innovators Dec [--] [----] https://t.co/I4PjNVea64 Yanggong Yifan Founder & CEO of 1st-gen [----] 2nd-gen (7nm) in [----] Local alternative to Google's TPU SCMP Nov [--] https://t.co/pthfwmLzVY https://t.co/4fVbKaZ2jQ VLIW AI Chip" [X Link](https://x.com/ogawa_tter/status/2006998357890507140) 2026-01-02T07:57Z [----] followers, [----] engagements "Tech Trends in Computing HW & their Impacts on High-Performance Sci Computing Part I: General-Purpose Processors & HW Accelerators UT Austin & ExxonMobil IJHPCA Dec [--] [----] (OnlineFirst) [--] pp [---] ref = Huawei Oct [----] https://x.com/ogawa_tter/status/2005532420033155255 https://journals.sagepub.com/doi/full/10.1177/10943420251348799 = Jul [--] [----] https://t.co/ZDi6UxGD7m "An Outlook for New Homogeneous Computing System" Huawei Fellow Oct [----] Unified Bus APNet [----] https://t.co/eYXz5ulaJU Sep [--] [----] https://t.co/LszYpCM0Xe UB-Mesh HC25 https://t.co/NXZRshSxio https://t.co/aZhtX280NL SIMT /" [X Link](https://x.com/ogawa_tter/status/2007084619737563357) 2026-01-02T13:40Z [----] followers, [---] engagements "= Baidu Announces Proposed Spin-off and Separate Listing of Kunlunxin () Jan [--] [----] Reuters valued it at $3B Dec [--] [----] Baidu World [----] Nov [--] https://x.com/ogawa_tter/status/1989203706043330700 https://www.reuters.com/world/asia-pacific/baidus-ai-chip-arm-kunlunxin-files-confidentially-hong-kong-listing-2026-01-01/ https://www.kunlunxin.com/ https://asia.cdn.euroland.com/press-releases-attachments./3955357/HKEX-EPS_20260102_11974277_0.PDF https://baidu.gcs-web.com/node/14381/pdf = (Baidu World) [----] Nov [--] https://t.co/0a4vhPeoMA Live https://t.co/L51mmUhB1V Kunlun () M100 (Early 2026)/M300" [X Link](https://x.com/ogawa_tter/status/2007656508474175726) 2026-01-04T03:33Z [----] followers, [---] engagements "= "KernelEvolve: Scaling Agentic Kernel Coding for Heterogeneous AI Accelerators at Meta" Dec [--] [----] arXiv (Jan [--] 2026) NVIDIA A100 / H100 AMD MI300 / MI350 MTIA v3 (Metas latest-gen AI accelerators) / v2i MTIA v2i ISCA [----] https://x.com/ogawa_tter/status/1936349816545763486 https://arxiv.org/abs/2512.23236 = "Meta's Second Generation AI Chip: Model-Chip Co-Design and Productionization Experiences" Industry Track ISCA [----] https://t.co/N34ujcpDh9 MTIA 2i Hot Chips [----] https://t.co/6xvaGCWAd9 https://t.co/m5Srr8wVXm Andes #RISCV Day Tokyo Feb [--] https://t.co/JWLJG79zKS Arteris" [X Link](https://x.com/ogawa_tter/status/2007741232240558427) 2026-01-04T09:09Z [----] followers, [---] engagements "= Biren Technology () One of China's "Four Little Dragons" of GPUs has entered the capital market Jan [--] [----] [----] HK GLOBAL OFFERING Dec [--] [----] (663 pp) Jul [--] [----] https://x.com/ogawa_tter/status/1950816787647570069 https://www1.hkexnews.hk/listedco/listconews/sehk/2025/1222/2025122200019.pdf https://www.hkex.com.hk/Market-Data/Securities-Prices/Equities/Equities-Quotesym=6082&sc_lang=zh-HK https://www.c114.com.cn/news/16/a1303403.html = "Communication Board Topology and Communication System" & (Xizhi Tech) Appl (CN: Feb 2025) https://t.co/t2f2upJhje GPUs are connected by Switch Chips/dOCSs" [X Link](https://x.com/ogawa_tter/status/2008108437931233495) 2026-01-05T09:28Z [----] followers, [---] engagements "= CXMT ()'s IPO application on the has been accepted Dec [--] [----] (Previewed) Prospectus Dec [--] (365 pp) no "HBM" word (Q&A) HBM 🤔🤔 https://static.sse.com.cn/stock/disclosure/announcement/c/202512/002170_20251230_XS2U.pdf https://static.sse.com.cn/stock/disclosure/announcement/c/202512/002170_20251230_B8QS.pdf https://www.sse.com.cn/listing/renewal/ipo/index_listing_detail.shtmlauditId=2170 https://eet-china.com/news/202512314093.html https://static.sse.com.cn/stock/disclosure/announcement/c/202512/002170_20251230_XS2U.pdf" [X Link](https://x.com/ogawa_tter/status/2008191584895176810) 2026-01-05T14:59Z [----] followers, [---] engagements "= "Hybrid Computing Network Topologies for Workload Execution" Brian P. Towles Yixin Luo Google (Mar 2024) can flexibly allocate bandwidth etc. "Enhanced Reconfigurable Interconnect Network" Granted: Jun [----] (P: Oct 2020) https://x.com/ogawa_tter/status/1619081874596585472 https://patents.google.com/patent/US20250284529A1/en = "Enhanced Reconfigurable Interconnect Network" Google Patent Appl Apr [----] (Dec 2020) https://t.co/fj1IEKREU1 Reconfigurable Interconnect Fabric for Scalability Optimized Network Bandwidth Optimized Network Latency Optimized Network TPUv4 Oct 2022" [X Link](https://x.com/ogawa_tter/status/2008456395105227008) 2026-01-06T08:31Z [----] followers, [---] engagements "@miyox @yyasuda @ebiken @O_FiberLady 9NVIDIA CEO/ Enfabrica Legal Events ENTITY STATUS OF PATENT OWNER : SMALL ENTITY 2025-10-30: LARGE ENTITY https://x.com/ogawa_tter/status/1620466846888775686 https://x.com/ogawa_tter/status/1968860255934173256 = Enfabrica (Jun 2022) Transparent Remote Memory Access ovr NW Protocol https://t.co/wRY8pBxT9O Multi-Plane /-Protocol Memory Switch Fabric w/ Configurable Transport https://t.co/aBB1cZ6WGa Server Fabric Adapter (Jan 2022) https://t.co/KU3T1lTSyv Jun [----] https://t.co/8rzxBuZnFU https://t.co/jfveOb2KGO" [X Link](https://x.com/ogawa_tter/status/2008493357199401099) 2026-01-06T10:58Z [----] followers, [---] engagements "= CES [----] Live with J. Huang Feb [--] Financial Analyst Q&A News Rubin - Six New Chips BF-4 Inference Context Memory Storage Platform https://nvidianews.nvidia.com/news/nvidia-bluefield-4-powers-new-class-of-ai-native-storage-infrastructure-for-the-next-frontier-of-ai https://nvidianews.nvidia.com/news/rubin-platform-ai-supercomputer https://nvidianews.nvidia.com/online-press-kit/ces-2026-news https://video.ibm.com/recorded/134654723 https://s201.q4cdn.com/141608511/files/doc_downloads/2026/01/05/JHH-CES-2026-FINAL.pdf https://www.youtube.com/watchv=0NBILspM4c4&t=6206s" [X Link](https://x.com/ogawa_tter/status/2008805340125819084) 2026-01-07T07:38Z [----] followers, [--] engagements "= Tenstorrent: First Gen Compact AI Accelerator Device Jan [--] [----] Partnership with Razer Wormhole tech Compact & Modular AI accelerator brings powerful GenAI capabilities to any Thunderbolt [--] or [--] equipped system up to [--] devices https://www.razer.com/newsroom/company-news/razer-at-ces-2026/ https://tenstorrent.com/vision/tenstorrent-unveils-first-gen-compact-ai-accelerator-device https://www.razer.com/newsroom/company-news/razer-at-ces-2026/ https://tenstorrent.com/vision/tenstorrent-unveils-first-gen-compact-ai-accelerator-device" [X Link](https://x.com/ogawa_tter/status/2008809454960341001) 2026-01-07T07:54Z [----] followers, [---] engagements "= AMD CES [----] Jan [--] Yottascale Era Press Kit Financial AD Nov [--] UALoE HPE Dec [--] Radiation Effects Oct [--] https://x.com/ogawa_tter/status/2001567210595324132 https://x.com/ogawa_tter/status/1996037944759746638 https://x.com/ogawa_tter/status/1988608702807691427 https://www.amd.com/en/newsroom/press-kits/ces-2026.html https://www.amd.com/en/blogs/2026/from-ces-2026-to-yottaflops-why-the-amd-keynote.html https://www.amd.com/content/dam/amd/en/documents/corporate/events/amd-ces-2026-distribution-deck.pdf https://www.youtube.com/watchv=UbfAhFxDomE "Radiation Effects in FPGAs and SoCs: Adapting" [X Link](https://x.com/ogawa_tter/status/2008824290058416253) 2026-01-07T08:53Z [----] followers, [---] engagements "= NVIDIA CES [----] J. Huang Feb [--] Financial Analyst Q&A News Rubin - Six New Chips BF-4 Inference Context Memory Storage Platform https://nvidianews.nvidia.com/news/nvidia-bluefield-4-powers-new-class-of-ai-native-storage-infrastructure-for-the-next-frontier-of-ai https://nvidianews.nvidia.com/news/rubin-platform-ai-supercomputer https://nvidianews.nvidia.com/online-press-kit/ces-2026-news https://video.ibm.com/recorded/134654723 https://s201.q4cdn.com/141608511/files/doc_downloads/2026/01/05/JHH-CES-2026-FINAL.pdf https://youtube.com/watchv=0NBILspM4c4&t=6206s" [X Link](https://x.com/ogawa_tter/status/2008827058689061146) 2026-01-07T09:04Z [----] followers, [--] engagements "= NVIDIA CES [----] J. Huang Feb [--] Financial Analyst Q&A News Rubin - Six New Chips BF-4 Inference Context Memory Storage Platform https://nvidianews.nvidia.com/news/nvidia-bluefield-4-powers-new-class-of-ai-native-storage-infrastructure-for-the-next-frontier-of-ai https://nvidianews.nvidia.com/news/rubin-platform-ai-supercomputer https://nvidianews.nvidia.com/online-press-kit/ces-2026-news https://video.ibm.com/recorded/134654723 https://s201.q4cdn.com/141608511/files/doc_downloads/2026/01/05/JHH-CES-2026-FINAL.pdf https://youtube.com/watchv=0NBILspM4c4&t=6206s" [X Link](https://x.com/ogawa_tter/status/2008828254308970667) 2026-01-07T09:09Z [----] followers, [---] engagements "= Intel CES [----] Core Ultra Series [--] Launch Event Jan [--] [----] Lip-Bu Tan Series [--] Deep Dive (88 pp) Press Kit Series [--] Debut as First Built on Intel 18A https://newsroom.intel.com/client-computing/ces-2026-intel-core-ultra-series-3-debut-first-built-on-intel-18a https://newsroom.intel.com/press-kit/press-kit-intel-at-ces-2026 https://download.intel.com/newsroom/2026/CES2026/Intel-CES2026-Core-Ultra-Series-3-Press-Deck.pdf https://download.intel.com/newsroom/2026/CES2026/Intel-CES%202026-Core-Ultra-Series-3-Keynote-010526.pdf https://www.youtube.com/watchv=KlIlFt2Fj1c" [X Link](https://x.com/ogawa_tter/status/2008861508885954844) 2026-01-07T11:21Z [----] followers, [---] engagements "= NVIDIA Inference Context Memory Storage Platform CES [----] Jan [--] BlueField-4 (Eth RDMA)-powered Flash-based G3.5 context memory tier Dynamo KVBM Oct [----] Enfabrica G2.5 Memory https://x.com/ogawa_tter/status/1968860255934173256 https://www.youtube.com/watchv=f5O3BrGwZVE https://developer.nvidia.com/blog/introducing-nvidia-bluefield-4-powered-inference-context-memory-storage-platform-for-the-next-frontier-of-ai/ https://x.com/ogawa_tter/status/2008828254308970667 = NVIDIA just spent over $900 million to hire @enfabrica_ CEO Rochan Sankar & other employees license AI startups technology Sep 18" [X Link](https://x.com/ogawa_tter/status/2009156839439777817) 2026-01-08T06:54Z [----] followers, [----] engagements "NVIDIA CES [----] Inference Context Memory Storage Platform BlueField-4 Eth RDMA Flash-based G3.5 context memory tier G1: GPU HBM G2: System RAM G3: Local SSDs G4: Shared storage Enfabrica G2.5 Mem Pool https://x.com/ogawa_tter/status/2009156839439777817 = NVIDIA Inference Context Memory Storage Platform CES [----] https://t.co/wMoR85lHT7 Jan [--] https://t.co/9GQsVVcXHp BlueField-4 (Eth RDMA)-powered Flash-based G3.5 context memory tier Dynamo KVBM Oct [----] https://t.co/UKOs1h0x64 Enfabrica https://t.co/jIGyT2BHx2 G2.5 Memory https://t.co/KwWwuvawvl" [X Link](https://x.com/ogawa_tter/status/2009166984253190577) 2026-01-08T07:35Z [----] followers, [---] engagements "= "Challenges and Research Directions for Large Language Model Inference Hardware" X. Ma David Patterson Google arXiv Jan [--] [----] [--] Research Opportunities to Re-Think HBF NMC 3D Stacking LL Interconnect https://x.com/ogawa_tter/status/1991789595214893165 https://x.com/ogawa_tter/status/1949068347565547896 https://arxiv.org/abs/2601.05047 = "Computer Architecture [---] & Its Future" David Patterson WS SimonsInst Sep [--] [----] https://t.co/pe4kVzqmHa TPU [--] yrs Aug [--] [----] https://t.co/fZJjK48sry [----] - [--] [----] https://t.co/MVZntG5mpc Emissions https://t.co/1y6H9cbs8Z How to Have a Bad" [X Link](https://x.com/ogawa_tter/status/2009473548679815209) 2026-01-09T03:53Z [----] followers, [---] engagements "= State-of-the-Art Analysis and Perspectives of China HPC Development in . [----] [----] [----] [----] [----] [----] http://www.jfdc.cnic.cn/CN/10.11871/jfdc.issn.2096-742X.2020.06.001 http://www.jfdc.cnic.cn/CN/10.11871/jfdc.10-1649.2021.06.007 http://www.jfdc.cnic.cn/CN/10.11871/jfdc.issn.2096-742X.2022.06.001 http://www.jfdc.cnic.cn/CN/10.11871/jfdc.issn.2096-742X.2023.06.001 http://www.jfdc.cnic.cn/CN/10.11871/jfdc.issn.2096-742X.2024.06.001 http://www.jfdc.cnic.cn/CN/10.11871/jfdc.issn.2096-742X.2025.06.009 http://www.jfdc.cnic.cn/CN/10.11871/jfdc.issn.2096-742X.2020.06.001" [X Link](https://x.com/ogawa_tter/status/2009623488596988004) 2026-01-09T13:49Z [----] followers, [---] engagements "= Tech Trends in Computing HW & their Impacts on High-Performance Sci Computing Part II: Memory systems Interconnects and System integration UT Austin & ExxonMobil IJHPCA Jan [--] [----] (OnlineFirst) [--] pp [---] ref Part I Dec [--] https://x.com/ogawa_tter/status/2007084619737563357 https://journals.sagepub.com/doi/full/10.1177/10943420251347461 Tech Trends in Computing HW & their Impacts on High-Performance Sci Computing Part I: General-Purpose Processors & HW Accelerators UT Austin & ExxonMobil IJHPCA Dec [--] [----] (OnlineFirst) https://t.co/aN24vurNQE [--] pp [---] ref = Huawei Oct 2021" [X Link](https://x.com/ogawa_tter/status/2009637922681901475) 2026-01-09T14:46Z [----] followers, [---] engagements "= Semiconductors: Information on Projects Funded to Strengthen U.S. Supply Chain Dec [--] [----] GAO cancels $45M Oregon State grant for Microfluidics May [--] terminates $285M contract for SMART USA Dec [--] https://x.com/ogawa_tter/status/2001156497288613985 https://x.com/ogawa_tter/status/1971941691994300867 https://www.gao.gov/products/gao-26-107882 = Federal gov terminates $285M contract for SMART USA CHIPS manufacturing institute Dec [--] [----] https://t.co/LeEYy4RuNm 'termination for convenience' Executive Director https://t.co/3ztuy4COqT ORAU Webinar Oct [--] https://t.co/hL1DxWApT3 Feb 19" [X Link](https://x.com/ogawa_tter/status/2009823914982256685) 2026-01-10T03:05Z [----] followers, [---] engagements "= "What is Groq-NVIDIA Deal Really About" Jan [--] [----] @sallywf "may include a desire to stop one of these hyperscale customers from buying Groq " Silicon Tiler Ecosystem Dense packag Deterministic HBM https://x.com/ogawa_tter/status/2004677049685008484 https://x.com/ogawa_tter/status/1770734477066608744 https://www.eetimes.com/what-is-groq-nvidia-deal-really-about/ = Groq & NVIDIA Enter Non-Exclusive Inference Tech Licensing Agreement Dec [--] [----] https://t.co/mGkLL3xafG J. Ross S. Madra & other members of the Groq team will join NVIDIA https://t.co/QqlSILw38M Dense packaging: HBM Stack" [X Link](https://x.com/ogawa_tter/status/2009903896161071440) 2026-01-10T08:23Z [----] followers, [---] engagements "= Xiaoyu Ma David Patterson. Google arXiv Jan [--] [----] "Tree-based Network Architecture for Accelerating Machine Learning Collective Operations" Nov [----] (Jan 2024) LLM with 3D-DRAM Chiplets (P: Jun 2023) https://x.com/ogawa_tter/status/1897205643519684768 https://patents.google.com/patent/US12481608B2/en https://x.com/ogawa_tter/status/2009473548679815209 = "Serving Large Language Models with 3D-DRAM Chiplets" Xiaoyu Ma David Patterson Google (P: Jun 2023) https://t.co/OjCeLg4HBd Compute-memory chiplets using small low-power comp dies with low arithmetic intensity (32 TF 40W TDP) NVIDIA" [X Link](https://x.com/ogawa_tter/status/2010264445755548026) 2026-01-11T08:16Z [----] followers, [---] engagements "= Interview w/ Mitesh Agarwal CEO of Positron Dec [--] [----] Journey to Redefine AI HW Two core innovations 90+% memory utilization Multi-model concurrency Taping out an ASIC Agilex-7M FPGA $51.6M Jul [--] https://x.com/ogawa_tter/status/1950213984264868004 https://x.com/ogawa_tter/status/1978738177457438975 https://eqvista.com/interview-with-mitesh-agarwal-positron/ = Positron AI Secures $51.6M in Oversubscribed Series A to Accelerate Inference-Optimized Hardware Jul [--] [----] https://t.co/GEJjxbaMt6 Atlas: First-gen FPGA-based $23.5M seed round Feb [--] https://t.co/cGsG7ssFiY Agilex-7M" [X Link](https://x.com/ogawa_tter/status/2010366527288262767) 2026-01-11T15:01Z [----] followers, [---] engagements "= "Performance efficiency and cost analysis of wafer-scale AI accelerators vs. single-chip GPUs" UC Riverside Review Device Oct [--] [----] Cerebras WSE-3 Tesla Dojo (SoW) NVIDIA H100 Google TPU v6e [--] pp [---] ref M. Ozkan https://scholar.google.com/citationsuser=zE75wgsAAAAJ&view_op=list_works&sortby=pubdate https://www.sciencedirect.com/science/article/pii/S2666998625001474 https://scholar.google.com/citationsuser=zE75wgsAAAAJ&view_op=list_works&sortby=pubdate https://www.sciencedirect.com/science/article/pii/S2666998625001474" [X Link](https://x.com/ogawa_tter/status/2010579494504132696) 2026-01-12T05:08Z [----] followers, [---] engagements "= IBM Z Tech Bytes Session Proceedings for May [----] May 15-15 IBM Z Performance Hot Topics (31 pp) Tech Overview (64 pp) Hot Chips [----] z17 Processor Module https://x.com/ogawa_tter/status/1918253608635888043 https://x.com/ogawa_tter/status/1911330163939291149 https://www.ibm.com/support/pages/system/files/inline-files/IBM%20z17%20and%20LinuxONE%205%20Technical%20Overview.pdf https://www.ibm.com/support/pages/system/files/inline-files/Tech%20Bytes%20May%202025%20IBM%20Z%20Hot%20Topics.pdf https://www.ibm.com/support/pages/node/7232338 = "The New IBM z17 Telum II Processor Module Cut Open Down" [X Link](https://x.com/ogawa_tter/status/2010994623079166033) 2026-01-13T08:37Z [----] followers, [---] engagements "= NN SoC in AR/VR glasses Meta Appls (Aug 2023) 3D Stacked Hybrid Memory MEMSYS [----] Stacked SRAM Face-to-Face Hybrid Bonding SRAM for AR Aug [----] (Sep 2021) https://patents.google.com/patent/US12068054B2/en https://patents.google.com/patent/US20250054910A1/en https://patents.google.com/patent/US20250054911A1/en https://x.com/ogawa_tter/status/1887752621361926464 https://patents.google.com/patent/US20250056815A1/en = "Characterization and Design of 3D-Stacked Memory for Image Signal Processing on AR/VR Devices" Meta MEMSYS [----] https://t.co/ZdzZAooBM5 3D-Stacked Hybrid Memory ISP solution" [X Link](https://x.com/ogawa_tter/status/2011057738282131777) 2026-01-13T12:48Z [----] followers, [---] engagements "= IBM Z Tech Bytes Session Proceedings for May [----] May 5-15 IBM Z Performance Hot Topics (31 pp) Tech Overview (64 pp) Hot Chips [----] z17 Processor Module https://x.com/ogawa_tter/status/1918253608635888043 https://x.com/ogawa_tter/status/1911330163939291149 https://www.ibm.com/support/pages/system/files/inline-files/IBM%20z17%20and%20LinuxONE%205%20Technical%20Overview.pdf https://www.ibm.com/support/pages/system/files/inline-files/Tech%20Bytes%20May%202025%20IBM%20Z%20Hot%20Topics.pdf https://www.ibm.com/support/pages/node/7232338 = "The New IBM z17 Telum II Processor Module Cut Open Down" [X Link](https://x.com/ogawa_tter/status/2011416904423141580) 2026-01-14T12:35Z [----] followers, [---] engagements "= "SmartBlock: Adaptive Block Floating Point Quantization for Efficient DNN Acceleration" NUDT ICPP [----] Sep [--] Systolic Array: Unified PE Arch Block Size: Dynamic Google (P: Apr 2024) Systolic Array Block Data Structured Sparse https://x.com/ogawa_tter/status/1997305240270618950 https://dl.acm.org/doi/10.1145/3754598.3754660 = "Structured Sparse Matrix Acceleration In Systolic Arrays " Google Appls (P: Apr 2024) https://t.co/4lEIYrhPPf Dense and Structured Fine-Grained Sparse Block Data Type TPUv7: Ironwood Hot Chips [----] https://t.co/gqG74DgNLB Self-Repairable Chip (Apr 2024) & BIST (Feb" [X Link](https://x.com/ogawa_tter/status/2011448137060335621) 2026-01-14T14:39Z [----] followers, [---] engagements "= "High-Performance Computing and Responsibly Reckless Algorithms" Jack Dongarra Keynote ICPP [----] Sep [--] "Published" China HPC Top [---] 2025-20 Ozaki Scheme NVIDIA SCA/HPCAsia [----] https://x.com/ogawa_tter/status/2000530683064938507 https://x.com/ogawa_tter/status/2009623488596988004 https://ssl.linklings.net/conferences/icpp/icpp2025_program/views/includes/files/evt_copy101s1.pdf = "Guaranteed DGEMM Accuracy While Using Reduced Precision Tensor Cores Through Extensions of the Ozaki Scheme" NVIDIA SCA/HPCAsia [----] Jan [--] https://t.co/KEN71PYZZP Ozaki-II scheme Dec [--] https://t.co/XqAeyqdVrr Oct" [X Link](https://x.com/ogawa_tter/status/2011458303285707154) 2026-01-14T15:20Z [----] followers, [---] engagements "= "OpenAI partners with Cerebras" Jan [--] [----] to add 750MW of ultra low-latency AI compute to our platform online in multiple tranches through [----] Wafer-Scale Systems Carbon Perspective Aug [----] Review Oct [----] [--] pp [---] ref https://x.com/ogawa_tter/status/2011071479715414082 https://openai.com/index/cerebras-partnership/ = "Wafer-Scale Systems: A Carbon Perspective" ACM SIGEnergy Energy Informatics Review Aug [--] [----] https://t.co/uIYXOkOtir Cerebras CS-3 vs H100 systems Performance Efficiency & Cost analysis of Wafer-Scale AI Accelerators vs. Single-Chip GPUs Oct [--] https://t.co/yRytINwLQH" [X Link](https://x.com/ogawa_tter/status/2011623853739688343) 2026-01-15T02:17Z [----] followers, [---] engagements "= AMD Appls (Jun 2024) "Balanced Latency Stacked Cache" [--] MB L2M cache: [--] cycle latency (= 14) "Apparatus and System for Multi-Purpose Macros for Vertical Die Interconnects" disclosed macro cells can be used in stacking caches https://patents.google.com/patent/US20250391815A1/en https://patents.google.com/patent/US20260003794A1/en https://patents.google.com/patent/US20250391815A1/en https://patents.google.com/patent/US20260003794A1/en" [X Link](https://x.com/ogawa_tter/status/2011695771067040216) 2026-01-15T07:03Z [----] followers, [---] engagements "= "A Path to 448Gbps" Tad Hofmeister Google Keynote OIF 448Gbps Signaling for AI WS Apr [--] [----] Report PAM6/8 are more realistic paths [---] G/Lane May [--] / Apr [--] CPC Sep [--] https://x.com/ogawa_tter/status/1976861738051289284 https://x.com/ogawa_tter/status/1930633833541742644 https://convergedigest.com/oif-keynote-google-urges-swift-move-to-448g-for-ai-clusters/ https://www.oiforum.com/wp-content/uploads/Tad-A-Path-to-448Gbps.pdf = "Revolutionizing Rackscale Connectivity using Co-Packaged Copper and Optics" Samtec & Marvell [----] ECOC Product Focus Sep [--] https://t.co/SAZM4HeUBE Marvell" [X Link](https://x.com/ogawa_tter/status/2011810407695163499) 2026-01-15T14:39Z [----] followers, [----] engagements "= "AI Chip Startup Etched Raises $500 Million to Take on NVIDIA" Jan [--] [----] Bloomberg according to people familiar with the matter Jun [--] [----] Page does not exist🤔 G. Uberti Co-Founder & CEO Feb [--] https://x.com/ogawa_tter/status/1805775421235151153 https://www.etched.com/ https://finance.yahoo.com/news/ai-chip-startup-etched-raises-205348019.html = "@Etched is Making the Biggest Bet in AI" Jun [--] [----] https://t.co/EOfmQAMeq1 Sohu the worlds first transformer ASIC can only run one algorithm One 8x Sohu server replaces 160x H100 GPUs. TSMC 4nm Gavin Uberti Co-Founder & CEO Etched Feb 28" [X Link](https://x.com/ogawa_tter/status/2011994333034266672) 2026-01-16T02:50Z [----] followers, [---] engagements "= "Trump administration sets GPU export rules that put Chinese buyers at the back of the queue" Jan [--] [----] Revision to License Review Policy for Advanced Computing Commodities Jan [--] covered by this rule are the NVIDIA H200 . https://www.federalregister.gov/documents/2026/01/15/2026-00789/revision-to-license-review-policy-for-advanced-computing-commodities https://www.theregister.com/2026/01/14/us_gpu_export_rules/ https://www.federalregister.gov/documents/2026/01/15/2026-00789/revision-to-license-review-policy-for-advanced-computing-commodities" [X Link](https://x.com/ogawa_tter/status/2012059526858719728) 2026-01-16T07:09Z [----] followers, [---] engagements "= Fact Sheet: Restoring American Semiconductor Manufacturing Leadership Through an Agreement on Trade & Investment with Taiwan Jan [--] [----] U.S. Dep of Commerce Direct Investments: $250+ billion Additional Investments: $250+ billion Industrial Parks https://www.commerce.gov/news/fact-sheets/2026/01/fact-sheet-restoring-american-semiconductor-manufacturing-leadership https://www.commerce.gov/news/fact-sheets/2026/01/fact-sheet-restoring-american-semiconductor-manufacturing-leadership" [X Link](https://x.com/ogawa_tter/status/2012112873942852039) 2026-01-16T10:41Z [----] followers, [---] engagements "= MatX (P: May 2024) Re-Rounding in Integrated Circuit for Variance Reduction in AI Ops R. Pope Acyclic Arch R. Pope & M. Gunter Mar [--] [----] $100M Series A https://x.com/reinerpope/status/1899486276233146530 https://x.com/ogawa_tter/status/1774422751253823814 https://patents.google.com/patent/US20250355715A1/en https://patents.google.com/patent/US20250355711A1/en https://matx.com/ MatX is designing chips and systems to 10x the computing power for the worlds largest AI workloads. Today we are pleased to announce the closing of a $100M Series A funding round led by @sparkcapital with" [X Link](https://x.com/ogawa_tter/status/2012152501349622085) 2026-01-16T13:18Z [----] followers, [---] engagements "= Breaking the Memory Wall: Scaling AI Inference with Innovative Memory Fanout Arch On-Demand (Nov [--] 2025) Credo Nov [--] OmniConnect 112G VSR Weaver Memory Fanout Gearbox M. Agrawal CEO Positron AI Dec [--] https://x.com/ogawa_tter/status/2010366527288262767 https://investors.credosemi.com/news-events/news/news-details/2025/Credo-Unveils-Industrys-First-Memory-Fanout-Gearbox-for-Scalable-High-Bandwidth-AI-Inference/default.aspx https://event.on24.com/wcc/r/5135492/10AE35BE7833F4F6B7C17C80FFFCA319 = Interview w/ Mitesh Agarwal CEO of Positron Dec [--] [----] https://t.co/32BRMNr4zG Journey to" [X Link](https://x.com/ogawa_tter/status/2012345610956108247) 2026-01-17T02:05Z [----] followers, [---] engagements "= "HyGain: High-performance Energy-efficient Hybrid Gain Cell-based Cache Hierarchy" ACM TACO Mar [--] [----] 2T Gain Cell Cache Memory OS GC Memory HSP Wong Stanford with TSMC NVIDIA IEDM [----] May [----] https://x.com/ogawa_tter/status/1994399791426703746 https://x.com/ogawa_tter/status/1996939609570529777 https://dl.acm.org/doi/10.1145/3572839 = (Oxide Semiconductor) Gain Cell Memory: Technology Integration & Memory Access H.-S. Philip Wong Stanford Stanford DAM Kickoff WS May [--] [----] https://t.co/kYNvY4ybUC Trans on Electron Devices Apr [--] https://t.co/BWIUxIO1l8 RAAAM Memory" [X Link](https://x.com/ogawa_tter/status/2012888601974145226) 2026-01-18T14:03Z [----] followers, [---] engagements "= "SK Hynix's $13B packaging facility promises more HBM for the AI bubble" Jan [--] [----] [----] Market Outlook "Focus on the HBM-led memory supercycle" Jan [--] SK HBM Market Share SK HBM4 Tech Leadership https://news.skhynix.co.kr/2026-market-outlook/ https://news.skhynix.co.kr/fact-02/ https://www.theregister.com/2026/01/13/sk_hbm_packaging/ https://news.skhynix.co.kr/2026-market-outlook/ https://news.skhynix.co.kr/fact-02/ https://www.theregister.com/2026/01/13/sk_hbm_packaging/" [X Link](https://x.com/ogawa_tter/status/2013168090058473533) 2026-01-19T08:34Z [----] followers, [---] engagements "= "Optimization and Benchmarking of Monolithically Stackable Gain Cell Memory for Last-Level Cache" GaTech & UV arXiv Mar [--] [----] 7/3 nm HD SRAM 64/128 MB AOS-based 2T-GC [---] MB Review Jan [----] Stanford TSMC NVIDIA: IEDM25 https://x.com/ogawa_tter/status/2010608213436289284 https://arxiv.org/abs/2503.06304 = "Oxide Semiconductor for Advanced Memory Architectures: Atomic Layer Deposition Key Requirement and Challenges" Hanyang Univ Review Nano-Micro Letters Jan [--] [----] https://t.co/qTyZdUtkyY [--] pp [---] ref Huawei Appls https://t.co/sgEIsl93RA Stanford https://t.co/DOugYZznRW" [X Link](https://x.com/ogawa_tter/status/2013262449739244018) 2026-01-19T14:49Z [----] followers, [----] engagements "= GPU: HBM + (LP) DDR DRAM NVIDIA Patent Oct [----] (Apr 2021) ASSYST [----] with GaTech H100 HBM + Tensor Offloading-Memory AMD "Integrated Memory Controller Circuitry" (P: Sep 2023) MI455X HBM4 + LPDDR5 https://x.com/SemiAnalysis_/status/2013310524109328580 https://patents.google.com/patent/US20250098181A1/en https://x.com/ogawa_tter/status/1722957240334946367 In AMD's CES [----] keynote we spotted [--] LPDDR5 DRAM packages beside the MI455XAMD has not said anything about this to date but SemiAnalysis has detailed this hybrid HBM4+LPDDR5 memory configuration in our article from last April. This is" [X Link](https://x.com/ogawa_tter/status/2013487091712245851) 2026-01-20T05:41Z [----] followers, [---] engagements "= "Micron breaks ground on humungous NY DRAM fab after beating bats and tree huggers" Jan [--] [----] had been bedeviled by environmental delays celebrates official groundbreaking at NY megafab site Dec [----] https://x.com/ogawa_tter/status/1866651828701761955 https://www.micron.com/us-expansion/new-york https://www.theregister.com/2026/01/16/micron_fab_breaks_ground/ = Dep of Commerce Awards CHIPS Incentives to Micron for Idaho & New York Projects and PMT for Virginia DRAM Project to Secure Domestic Supply of Legacy Memory Dec [--] [----] https://t.co/Y0EDFzrfcm $4.6B & $1.5B in Direct + $65M VA PMT:" [X Link](https://x.com/ogawa_tter/status/2013505716972384486) 2026-01-20T06:55Z [----] followers, [---] engagements "@1p_semicon (4/8 Gain Cell Memory Stanford H.-S. Philip Wong TSMC 2018-20 VP Corporate Res Chief Scientist IEDM [----] NVIDIA GC-Memory RT https://x.com/ogawa_tter/status/1996939609570529777 = Gain Cell Memory IEDM [----] Dec [--] H.-S. Philip Wong Stanford DAM https://t.co/jNi5fUFjbv https://t.co/Zyrfs3WAY9 & TSMC NVIDIA 28-8 "Amp Cell: 3D-Scalable Gain Cell Memory with Decoupled Read Write and Store" 29-5 "Gain Cell Memory Scalability to 5-nm and Beyond" +Purdue https://t.co/Y5SZaAKP15 https://x.com/ogawa_tter/status/1996939609570529777 = Gain Cell Memory IEDM [----] Dec [--] H.-S. Philip Wong" [X Link](https://x.com/ogawa_tter/status/2013846225938981188) 2026-01-21T05:28Z [----] followers, [----] engagements "@1p_semicon (5/8 Preprint (arXiv Mar [--] 2025) Amorphous Oxide Semiconductor-based 2-Transistor Gain Cell Memory (AOS-based 2T-GC Memory) High-Density SRAMLast-Level Cache (64/128 MB) 7nm / 3nm https://x.com/ogawa_tter/status/2013262449739244018 = "Optimization and Benchmarking of Monolithically Stackable Gain Cell Memory for Last-Level Cache" GaTech & UV arXiv Mar [--] [----] https://t.co/kxph41Y1RE 7/3 nm HD SRAM 64/128 MB AOS-based 2T-GC [---] MB Review Jan [----] https://t.co/0szYhdisDf Stanford TSMC NVIDIA: IEDM25 https://t.co/rvxaShp0yf https://x.com/ogawa_tter/status/2013262449739244018 =" [X Link](https://x.com/ogawa_tter/status/2013846446957830365) 2026-01-21T05:29Z [----] followers, [----] engagements "@1p_semicon (6/8 Amorphous Oxide Semicon-based Tr2 H.-S. Philip Wong Oxide Semicon OS for Advanced Memory Arch Review (Nano-Micro Letters Jan [--] 2026) IGZO Memory https://x.com/ogawa_tter/status/2010608213436289284 = "Oxide Semiconductor for Advanced Memory Architectures: Atomic Layer Deposition Key Requirement and Challenges" Hanyang Univ Review Nano-Micro Letters Jan [--] [----] https://t.co/qTyZdUtkyY [--] pp [---] ref Huawei Appls https://t.co/sgEIsl93RA Stanford https://t.co/DOugYZznRW https://t.co/wJDqGcUgxR https://x.com/ogawa_tter/status/2010608213436289284 = "Oxide Semiconductor for Advanced" [X Link](https://x.com/ogawa_tter/status/2013846714793251064) 2026-01-21T05:30Z [----] followers, [----] engagements "= Upscale AI Closes Oversubscribed $200M Series A Jan [--] [----] Tota: $300+M Oct [--] [----] $100M Seed Sep [--] SkyHammer Arch Oct [--] Aug [----] (Apr 2025) to separate Header and Data https://patents.google.com/patent/US12388770B1/en https://upscaleai.com/upscale-ai-unveils-skyhammer-architecture/ https://x.com/ogawa_tter/status/1981691384538316808 https://upscaleai.com/from-100m-seed-to-unicorn-in-months-upscale-ai-closes-oversubscribed-200m-series-a-to-build-the-first-pure-play-ai-networking-company/ = Perform Evaluation of Interconnect Tech for AI Scale-Up Comp: UAL vs UALoE/SUE vs RoCE Upscale AI" [X Link](https://x.com/ogawa_tter/status/2014173579886068056) 2026-01-22T03:09Z [----] followers, [---] engagements "= "VCSEL-based CPO for Scale-Up in A.I. Datacenter. Status and Perspectives" arXiv Jan [--] [----] Mature VCSEL offers the ideal combination M Kohli COHR J Teissier COHR VCSEL COHR https://www.coherent.com/networking/optoelectronic-devices/datacom-vcsel-photodiodes/vcsel-arrays https://scholar.google.com/citationsuser=BeW1B9wAAAAJ&view_op=list_works&sortby=pubdate https://scholar.google.com/citationsuser=kMPss_sAAAAJ&view_op=list_works&sortby=pubdate https://arxiv.org/abs/2601.14342 https://www.coherent.com/networking/optoelectronic-devices/datacom-vcsel-photodiodes/vcsel-arrays" [X Link](https://x.com/ogawa_tter/status/2014212480340578694) 2026-01-22T05:44Z [----] followers, [---] engagements "@kazunori_279 @kumagi @miyox @yyasuda @ebiken @O_FiberLady ( HBF (High Bandwidth Flash) Jul [--] [----] Patterson Sandisk HBF Technical Advisory Board arXiv LLM 3D Stack Chips ML Collective Operations () ( https://x.com/ogawa_tter/status/2010264445755548026 https://x.com/ogawa_tter/status/1949068347565547896 = Xiaoyu Ma David Patterson. Google arXiv Jan [--] [----] https://t.co/lvUhAzkOqo "Tree-based Network Architecture for Accelerating Machine Learning Collective Operations" Nov [----] (Jan 2024) https://t.co/R5vQkMJY1j LLM with 3D-DRAM Chiplets (P: Jun 2023) https://t.co/GURhSDp8uu" [X Link](https://x.com/ogawa_tter/status/2014578608287957413) 2026-01-23T05:59Z [----] followers, [---] engagements "= "How to Write High-Performance Matrix Multiply in NVIDIA CUDA Tile" Jan [--] [----] How to implement matrix multiplication using NVIDIA cuTile Block-level parallel programming mindset Best practices GPU Mode Dec [--] [----] (1:43:20) https://x.com/ogawa_tter/status/2002971559053639796 https://developer.nvidia.com/blog/how-to-write-high-performance-matrix-multiply-in-nvidia-cuda-tile/ = NVIDIA CUDA Tile using cuTile and CUDA Tile IR to write portable high-performance GPU Kernels @JokerEph & @roeschinc GPU Mode Dec [--] [----] https://t.co/yAqJbFs1OW https://t.co/nuRPJUHJ8m cuTile Python Dec 4" [X Link](https://x.com/ogawa_tter/status/2015251010999222772) 2026-01-25T02:30Z [----] followers, [---] engagements "@t3840 "Wafer-Scale Systems: A Carbon Perspective" Harvard & FAIR at Meta ACM SIGEnergy Energy Informatics Review Aug [--] [----] LLMs CS-3 H100 Wafer-Scale AI Accelerators vs. Single-Chip GPUs Review (33 pp [---] ref) https://x.com/ogawa_tter/status/2011071479715414082 = "Wafer-Scale Systems: A Carbon Perspective" ACM SIGEnergy Energy Informatics Review Aug [--] [----] https://t.co/uIYXOkOtir Cerebras CS-3 vs H100 systems Performance Efficiency & Cost analysis of Wafer-Scale AI Accelerators vs. Single-Chip GPUs Oct [--] https://t.co/yRytINwLQH https://t.co/XpP730P2rN" [X Link](https://x.com/ogawa_tter/status/2015378525398991073) 2026-01-25T10:57Z [----] followers, [---] engagements "= "Cloud-Hardware Co-Design for Memory Bandwidth-Bound HPC Workloads: Performance and Characteristics of Azure HBv5 Virtual Machines" Microsoft & AMD SCA/HPCAsia [----] Jan [--] Custom-designed AMD EPYC 9V64H 8x HBM3 Nov [--] [----] https://x.com/ogawa_tter/status/1859053986198040579 https://dl.acm.org/doi/10.1145/3773656.3773684 = Azure HBv5 Virtual Machines: A Breakthrough in Memory Bandwidth for HPC Nov [--] [----] https://t.co/oMEDMRo4Jt 4x Custom 4th Gen EPYC with HBM3 Memory Bandwidth: [---] TB/s 400-450 GB Up to [---] Zen4 cores SMT disabled [---] Gb/s IB [---] Gb/s per CPU SoC [---] Gb/s: Azure Boost NIC" [X Link](https://x.com/ogawa_tter/status/2015715361111609355) 2026-01-26T09:16Z [----] followers, [---] engagements "Announcing Hortonworks Data Platform (HDP) version [---] GA: Enterprise Hadoop Apr [--] [----] http://t.co/TGG7X2oTOE http://hortonworks.com/blog/announcing-hdp-2-1-general-availability/ http://hortonworks.com/blog/announcing-hdp-2-1-general-availability/" [X Link](https://x.com/ogawa_tter/status/458917982895218690) 2014-04-23T10:39Z [----] followers, [--] engagements ""Common LISP Modules Artificial Intelligence in the Era of Neural Networks and Chaos Theory" M. Watson SAIC [----] http://link.springer.com/book/10.1007/978-1-4612-3186-8 http://link.springer.com/book/10.1007/978-1-4612-3186-8" [X Link](https://x.com/ogawa_tter/status/680959378879549440) 2015-12-27T03:52Z [----] followers, [--] engagements ".@ogawa_tter "The Quantum Software Challenge" What SW Engineers should know about Quantum Computing Oct [--] [----] https://medium.com/@rigetticomputing/the-quantum-software-challenge-1a86eec1ce47 https://medium.com/@rigetticomputing/the-quantum-software-challenge-1a86eec1ce47" [X Link](https://x.com/ogawa_tter/status/698511683375869953) 2016-02-13T14:18Z [----] followers, [--] engagements ".@ogawa_tter Rigetti Computing About 2016/2/13 [--] PhD12 Google Scholar8 arXiv6 MS2 Operations Manager http://rigetti.com/about http://rigetti.com/about" [X Link](https://x.com/ogawa_tter/status/698564703635673088) 2016-02-13T17:49Z [----] followers, [--] engagements ".@ogawa_tter Rigetti Computing Feb [--] [----] "able to accelerate some kinds of machine learning and run highly accurate chemistry simulations"" [X Link](https://x.com/ogawa_tter/status/700243861881237505) 2016-02-18T09:02Z [----] followers, [--] engagements ".@ogawa_tter A Practical Quantum Instruction Set Architecture Rigetti Comp Aug [--] [----] https://arxiv.org/abs/1608.03355 https://twitter.com/ogawa_tter/status/764072652214374400 https://arxiv.org/abs/1608.03355 https://twitter.com/ogawa_tter/status/764072652214374400" [X Link](https://x.com/ogawa_tter/status/788040462988775424) 2016-10-17T15:34Z [----] followers, [--] engagements ".@ogawa_tter Emerson W. Pugh Building IBM IBM's [---] and Early [---] Systems Oral History [----] http://ethw.org/Oral-History:Emerson_W._Pugh https://mitpress.mit.edu/authors/emerson-w-pugh http://ethw.org/Oral-History:Emerson_W._Pugh https://mitpress.mit.edu/authors/emerson-w-pugh" [X Link](https://x.com/ogawa_tter/status/820452918348288001) 2017-01-15T02:09Z [----] followers, [--] engagements ".@ogawa_tter IBM System/360 Model [--] Ref P. [--] Jan [----] PDF https://en.wikipedia.org/wiki/IBM_System/360_Model_85 http://bitsavers.trailing-edge.com/pdf/ibm/360/funcChar/A22-6916-1_360-85_funcChar_Jun68.pdf http://www.techrepublic.com/pictures/photos-looking-back-to-the-birth-of-the-mainframe/2/ https://en.wikipedia.org/wiki/IBM_System/360_Model_85 http://bitsavers.trailing-edge.com/pdf/ibm/360/funcChar/A22-6916-1_360-85_funcChar_Jun68.pdf http://www.techrepublic.com/pictures/photos-looking-back-to-the-birth-of-the-mainframe/2/" [X Link](https://x.com/ogawa_tter/status/820461414535995392) 2017-01-15T02:43Z [----] followers, [--] engagements "= "The DOE Exascale Computing Project (ECP): Overview of Relevant Energy/Power Efforts" EE HPC WG Nov [--] [----] ECP ST Software Stack ST: Runtime System for Application-Level Power Steering Ref: Sep [--] [----] https://x.com/ogawa_tter/status/918089367624630272 https://eehpcwg.llnl.gov/documents/conference/sc17/SC17_workshop_1530_exascale_ang.pdf = Exascale Computing Project Update P. Messina & S. Lee DoE ASCAC Sep [--] [----] PDF https://t.co/CgJxCEzA4G Changes in tech areas https://t.co/y3vj0BvNoD https://x.com/ogawa_tter/status/918089367624630272" [X Link](https://x.com/ogawa_tter/status/935083279664820224) 2017-11-27T09:50Z [----] followers, [--] engagements "= Oak Ridge Leadership Computing Facility: Recent updates and roadmap forward ASCR BigPanDA SC17 UTA booth Nov [--] [----] PDF Frontier: The OLCF [----] Exascale computer ORNL [----] Summit https://indico.cern.ch/event/680965/contributions/2790231/attachments/1558510/2452141/Wells_UT-ABooth_SC17.pdf https://indico.cern.ch/event/680965/contributions/2790231/attachments/1558510/2452141/Wells_UT-ABooth_SC17.pdf" [X Link](https://x.com/ogawa_tter/status/940213972027916288) 2017-12-11T13:37Z [----] followers, [--] engagements "= Oak Ridge Leadership Computing Facility: Recent updates and roadmap forward ASCR BigPanDA SC17 UTA booth Nov [--] (NOT 04) [----] Frontier: The OLCF [----] Exascale computer ORNL [----] https://x.com/ogawa_tter/status/940213972027916288 = Oak Ridge Leadership Computing Facility: Recent updates and roadmap forward ASCR BigPanDA SC17 UTA booth Nov [--] [----] PDF https://t.co/bT0lFnNobQ Frontier: The OLCF [----] Exascale computer ORNL [----] Summit https://t.co/Uzx9AbYVtz https://x.com/ogawa_tter/status/940213972027916288 = Oak Ridge Leadership Computing Facility: Recent updates and roadmap forward ASCR" [X Link](https://x.com/ogawa_tter/status/940225747704201216) 2017-12-11T14:24Z [----] followers, [--] engagements "Rigetti Blake Johnson HPC User Forum Apr [--] [----] PDF Forest An API for quantum computing in the cloud Sep [--] [----] http://www.blakerobertjohnson.com/about https://www.nature.com/news/first-quantum-computers-need-smart-software-1.22590 https://www.rigetti.com/forest https://rigetti.com/ https://www.hpcuserforum.com/presentations/tuscon2018/QCOverview_Rigetti_UFTucson2018.pdf https://www.youtube.com/watchv=VutplvAZxNw http://www.blakerobertjohnson.com/about https://www.nature.com/news/first-quantum-computers-need-smart-software-1.22590 https://www.rigetti.com/forest https://rigetti.com/" [X Link](https://x.com/ogawa_tter/status/989566936034885632) 2018-04-26T18:08Z [----] followers, [--] engagements "= Quantum Computing Sessions HPC User Forum Apr [--] [----] 2/4 J. Gambetta IBM S. Reinhardt D-Wave PDF Slides D-Wave Qubits Europe users conf Apr 11-12 [----] = https://x.com/ogawa_tter/status/988627912495529984 https://www.hpcuserforum.com/presentations/tuscon2018/QCOverview_DWave_UFTucson2018.pdf https://www.youtube.com/watchv=4s9vEaVJjdA https://www.youtube.com/watchv=TLE6ypLgIo0 = Slides [----] D-Wave Qubits Europe users conference Apr 11-12 https://t.co/afQGcOpPjP D-Wave = Next-gen architecture: Pegasus P6 https://t.co/UqYN55jiTM https://x.com/ogawa_tter/status/988627912495529984" [X Link](https://x.com/ogawa_tter/status/989576030166827009) 2018-04-26T18:44Z [----] followers, [--] engagements "= Quantum Computing Sessions HPC User Forum Apr [--] [----] 3/4 M. Troyer MSR B. Johnson Rigetti C. Williams NIST Zhu CAS CoE Chris Monroe IonQ = https://ionq.co/ https://x.com/ogawa_tter/status/989554968246927360 https://x.com/ogawa_tter/status/989556678000496640 https://x.com/ogawa_tter/status/989566936034885632 https://x.com/ogawa_tter/status/989561091737899008 "High Performance Quantum Computing" Matthias Troyer Microsoft Research HPC User Forum Apr [--] [----] https://t.co/TCN7t7CAEh PDF https://t.co/DjkWMM1ChQ https://t.co/CJCbp37v1m QC at MSR https://t.co/Ui4nWxROOf Microsoft Quantum" [X Link](https://x.com/ogawa_tter/status/989576328100769792) 2018-04-26T18:45Z [----] followers, [--] engagements Limited data mode. Full metrics available with subscription: lunarcrush.com/pricing
@ogawa_tter OGAWA, Tadashi
OGAWA, Tadashi posts on X about ai, dram, $googl, huawei the most. They currently have [-----] followers and [---] posts still getting attention that total [-----] engagements in the last [--] hours.
Engagements: [-----] #
- [--] Week [------] +45%
- [--] Month [------] +61%
- [--] Months [-------] -17%
- [--] Year [---------] -31%
Mentions: [--] #
- [--] Week [--] +120%
- [--] Month [---] -64%
- [--] Months [---] +64%
- [--] Year [-----] +16%
Followers: [-----] #
- [--] Week [-----] -0.07%
- [--] Month [-----] +0.33%
- [--] Months [-----] +2.90%
- [--] Year [-----] +9.40%
CreatorRank: [---------] #
Social Influence
Social category influence technology brands 54.84% stocks 47.31% currencies 11.29% finance 8.06% countries 5.91% automotive brands 2.15% travel destinations 1.61% events 1.61% cryptocurrencies 1.61% vc firms 0.54%
Social topic influence ai 24.19%, dram #178, $googl 9.68%, huawei 8.6%, dec #270, inference 6.99%, systems 5.91%, ceo 5.38%, core 4.84%, ibm 4.3%
Top accounts mentioned or mentioned by @1psemicon @ogawatter @onurmutlu @yyasuda @semianalysis @jokereph @roeschinc @miyox @ebiken @ofiberlady @50 @inouefumihiro @suekichiii @cqcqcqdx @bcantrill @instlatx64 @yuya0240 @akigoodspeed @zenrinofficial @mak0nakamura
Top assets mentioned Alphabet Inc Class A (GOOGL) IBM (IBM) Microsoft Corp. (MSFT) IonQ, Inc. (IONQ) Tesla, Inc. (TSLA) Exxon Mobil (XOM) SkyWater Technology, Inc. (SKYT) GlobalFoundries (GFS)
Top Social Posts
Top posts by engagements in the last [--] hours
"= HUAWEI CONNECT [----] Sep [--] Ascend 950PR (Prefill & Recommend) [----] Q1 950DT (Decode & Training) [----] Q4 [--] PF (FP8)/2 PF (FP4) FP32/HF32/FP16/BF16/FP8/MXFP8/HiF8/MXFP4 Proprietary HBM NVIDIA Rubin CPX Disaggregated https://x.com/ogawa_tter/status/1966342884363497979 https://www.ithome.com/0/883/839.htm = Another Giant Leap: The Rubin CPX Specialized Accelerator & Rack Sep [--] [----] @SemiAnalysis_ https://t.co/RkmMlYxAfV The Giant Leap: Disaggregated Serving lan Buck Sep [--] https://t.co/nFfaV9zR5e Bill Dally Aug [--] https://t.co/ksqVYCR6A5 https://t.co/ssxBrbz3WG Disaggregated"
X Link 2025-09-18T21:20Z [----] followers, [----] engagements
"= Research Progress of High-Degree OXC Technologies Invited Jul [--] [----] Huawei OptiXtrans DC808 Sep [--] [----] Special Prize at Interop Tokyo [----] Jun [--] Beyond CLOS https://x.com/ogawa_tter/status/1976552928443035975 https://e.huawei.com/en/news/2025/solutions/enterprise-optical-network/data-center-all-optical-switch https://e.huawei.com/cn/news/2024/industries/grid/all-optical-switch-optixtrans-dc808 https://www.opticsjournal.net/Articles/OJbc1ca6bc13ed3246/Abstract State of the Opt Components Industry Omdia ECOC [----] Market Focus https://t.co/BRNpAGZdha = Beyond CLOS Futurewei ECOC 2025"
X Link 2025-11-04T03:30Z [----] followers, [----] engagements
"= "IonQ to Acquire SkyWater Technology Creating the Only Vertically Integrated Full-Stack Quantum Platform Company" Jan [--] [----] $1.8B ($35.00/share) SkyWater Jan [--] Strong momentum in quantum computing. https://s27.q4cdn.com/656702482/files/doc_events/2026/Jan/14/SKYT-IR-Presentation-for-NGC-Jan-2026.pdf https://s28.q4cdn.com/828571518/files/doc_downloads/2026/01/26/IonQ-to-Acquire-SkyWater-Technology-2026-01-26-vFF.pdf https://investors.ionq.com/news/news-details/2026/IonQ-to-Acquire-SkyWater-Technology-Creating-the-Only-Vertically-Integrated-Full-Stack-Quantum-Platform-Company/default.aspx"
X Link 2026-01-26T14:14Z [----] followers, [---] engagements
"= "Tower Semiconductor Teams with NVIDIA to Advance AI Infrastructure with 1.6T Data Center Optical Modules" Feb [--] [----] Silicon photonics for 1.6T data center optical modules designed for NVIDIA networking protocols SiPho https://towersemi.com/technology/rf-and-hpa/silicon-photonics-rf/ https://towersemi.com/2026/02/05/02052026/ https://towersemi.com/technology/rf-and-hpa/silicon-photonics-rf/ https://towersemi.com/2026/02/05/02052026/"
X Link 2026-02-07T03:56Z [----] followers, [---] engagements
"= Select Committee on the CCP REPORT: American VC Firms Investing Billions into PRC Companies Fueling the CCP's Military Surveillance State and Uyghur Genocide Feb [--] [----] GGV Capital GSR Ventures Qualcomm Ventures Sequoia Capital (China) Walden Inter https://selectcommitteeontheccp.house.gov/media/press-releases/committee-report-american-vc-firms-investing-billions-prc-companies-fueling https://selectcommitteeontheccp.house.gov/media/press-releases/committee-report-american-vc-firms-investing-billions-prc-companies-fueling"
X Link 2024-02-09T09:27Z [----] followers, 20.3K engagements
"= William James Dally [----] Benjamin Franklin Medal in Computer & Cognitive Science Merrimac SC03 Brook for GPUs SIGGRAPH [----] Chris Malachowsky NVIDIA EDPS [----] Stream Processors ACM Queu Mar [----] https://x.com/ogawa_tter/status/1611750717865463808 https://x.com/ogawa_tter/status/1733150189450412442 https://fi.edu/en/awards/laureates/william-james-dally = Stream Processors Mar [----] https://t.co/SG9FO9aHAm SIGRAPH [----] CUDA Jun [----] https://t.co/wNi9fSqDDI Scalable Parallel Programming with CUDA Mar/Apr [----] https://t.co/9PapnzQYMs The GPU Computing Era Micro Mar/Apr 2010"
X Link 2025-11-08T13:03Z [----] followers, [----] engagements
""Government University & Industry Cooperation: The NVIDIA Story" Bill Dally NSB Meeting Jul [--] [----] D. Kirk SC19 IAA WS [----] = Benjamin Franklin Medal https://x.com/ogawa_tter/status/1987143958276059280 https://x.com/ogawa_tter/status/1599768199759040513 https://x.com/ogawa_tter/status/1205627606533464064 https://nsf-gov-resources.nsf.gov/files/SPEAKER_2_BDally_panel_slides.pdf https://www.youtube.com/live/9Os9sAAcsjMt=842s = William James Dally [----] Benjamin Franklin Medal in Computer & Cognitive Science https://t.co/OtFWa4CqZP https://t.co/WucQGgalko Merrimac SC03 Brook for GPUs SIGGRAPH"
X Link 2025-11-17T14:12Z [----] followers, [----] engagements
"= "Addressing Memory Contention in Modern DRAM-Based Systems" M. Patel Rutgers SAFARI Conf Dec [--] [----] PhD Thesis [----] @onurmutlu R/W in parallel POWER9=Power10 Google Appl https://x.com/ogawa_tter/status/1896529002199970293 https://x.com/ogawa_tter/status/1879535899350085707 https://events.safari.ethz.ch/SafariConference25/lib/exe/fetch.phpmedia=minesh-safari-conference-contention-afterlecture.pptx https://www.youtube.com/live/QtS7v8aVGLQt=713s = Google Systems & Methods for HBM with Unidirectional Data Flow (P: Aug 2023) https://t.co/aCwOci14fs Concurrent Read & Write in parallel HBM4"
X Link 2025-12-21T14:13Z [----] followers, [----] engagements
"= "Efficient Kernel Mapping and Comprehensive System Evaluation of LLM Acceleration on a CGLA" T. Ando . Y. Nakashima ( ) NAIST (LENZO) IEEE Access Dec [--] [----] LENZO [-------] [----] IMAX (CGLA) NAIST https://x.com/ogawa_tter/status/1987877170291876102 https://ieeexplore.ieee.org/document/11264489 CGLANVIDIA2025827 https://t.co/Qz7I42l8BG LENZO NVIDIA = [----] https://t.co/EAoQ5QlZJW CGRA vs CGLA Jul [--] VPP500 ICS '95 IMAX (CGLA) NAIST https://t.co/DpUKXPUyIr https://t.co/Y6hOwEf3CO https://x.com/ogawa_tter/status/1987877170291876102 https://ieeexplore.ieee.org/document/11264489 CGLANVIDIA2025827"
X Link 2026-01-04T15:08Z [----] followers, [----] engagements
"NVIDIA CGLA202611 "Architecture Leads System Design" Y. Nakashima NAIST/LENZO Invited Nov [--] [----] = IEEE Access Dec [--] [----] FPGA Prototype https://x.com/ogawa_tter/status/2007831406567903258 https://www.youtube.com/watchv=gV90-tfVTbg https://www.youtube.com/watchv=1LBNnC37CKI = "Efficient Kernel Mapping and Comprehensive System Evaluation of LLM Acceleration on a CGLA" T. Ando . Y. Nakashima ( ) NAIST (LENZO) IEEE Access Dec [--] [----] https://t.co/qflvzQnf3H LENZO [-------] https://t.co/79Ogr7aMHV [----] IMAX (CGLA) NAIST https://t.co/EWlm1GI8i2 https://x.com/ogawa_tter/status/2007831406567903258"
X Link 2026-01-05T02:07Z [----] followers, [----] engagements
"= Advances and Future Trends in Photolithography & Photoresist Materials Invited Apr [----] [--] pp [---] ref Research Progress & Prospects of Extreme Ultraviolet Photoresists Sep [----] May [----] https://x.com/ogawa_tter/status/1653573723746881536 https://www.opticsjournal.net/Articles/OJf78e7687596f5910/Abstract https://www.opticsjournal.net/Articles/OJ762ec4ae17d37064/Abstract = [-----] EUV Lithography Source https://t.co/M8succaWMF Stage Control Methods https://t.co/7a3TXtSyOx EUV Photoresists https://t.co/ON84ttXsLy Computational Lithography https://t.co/g8bGJhxFTm Advanced Photolithography"
X Link 2026-01-14T12:48Z [----] followers, [---] engagements
"= GlobalFoundriestoAcquireSynopsysProcessor IPSolutions Business ExpandingCapabilitiesto Accelerate Physical AIApplications Jan [--] [----] ARC-V ARC-Classic ARCVPX-DSPandARCNPX NPU ASIP tools GF: Physical AI Dec [--] https://investors.gf.com/events/event-details/gf-business-webinar-series-gf-enabling-evolution-edge-physical-ai https://gf.com/gf-press-release/globalfoundries-to-acquire-synopsys-processor-ip-solutions-business/ https://investors.gf.com/events/event-details/gf-business-webinar-series-gf-enabling-evolution-edge-physical-ai"
X Link 2026-01-15T02:45Z [----] followers, [---] engagements
"= "SiFive to Power Next-Gen #RISCV AI Data Centers with NVIDIA NVLink Fusion" Jan [--] [----] NVLink Fusion HotI [----] https://x.com/ogawa_tter/status/1962326039646097646 https://www.nvidia.com/en-us/data-center/nvlink-fusion/ https://www.theregister.com/2026/01/15/sifive_nvidia_nvlink/ https://www.sifive.com/press/sifive-nvidia-nvlinkfusion-datacenter = Building Custom AI Infrastructure w/ NVLink Fusion K Geeyarpuram Invited HotI [----] Aug [--] https://t.co/nazZi87SOl https://t.co/lMJHANLJag NVLink & NVLink Fusion Aug [--] https://t.co/uLmE6tSDkH "FugakuNEXT" Aug [--] https://t.co/FJdsUNdhIc Fujitsu"
X Link 2026-01-16T03:10Z [----] followers, [---] engagements
"= "Synopsys Enters Definitive Agreement with GlobalFoundries For Sale of Processor IP Solutions Business" Jan [--] [----] Synopsys Processor Solutions ARC-V (#RISCV) etc GF Physical AI https://x.com/ogawa_tter/status/2011630756267987059 https://www.synopsys.com/dw/doc.php/ds/cc/dw-processor-solutions.pdf https://www.synopsys.com/designware-ip/processor-solutions.html https://investor.synopsys.com/news/news-details/2026/Synopsys-Enters-Definitive-Agreement-with-GlobalFoundries-For-Sale-of-Processor-IP-Solutions-Business/default.aspx = GlobalFoundriestoAcquireSynopsysProcessor IPSolutions Business"
X Link 2026-01-16T06:33Z [----] followers, [---] engagements
"= IBM Buffered Server Memory POWER9 Power10: [----] GB/s OMI DDR4 Power11 Jul [----] [----] GB/s OMI DDR5 [--] TB [----] GB/s/Skt Credo OmniConnect Core AXI over 112G VSR = Weaver: LPDDR Positron Asimov N3P ASIC https://x.com/ogawa_tter/status/2012345610956108247 https://credosemi.com/products/omniconnect/ https://x.com/ogawa_tter/status/1950024476684419076 = Breaking the Memory Wall: Scaling AI Inference with Innovative Memory Fanout Arch On-Demand (Nov [--] 2025) https://t.co/nHCZ0RuB8N Credo Nov [--] https://t.co/468eqr6Ol8 OmniConnect 112G VSR Weaver Memory Fanout Gearbox M. Agrawal CEO Positron AI Dec"
X Link 2026-01-17T04:37Z [----] followers, [----] engagements
"= Google Appls (PCT: Mar 2025) Memory & Compute 3DIC Integration without TSV Thermal Management for Integrated Memory & Compute Circuits X. Ma & D. Patterson arXiv Jan [--] (8) [----] Processing-Near-Memory https://x.com/ogawa_tter/status/2010264445755548026 https://patents.google.com/patent/WO2025189071A1/en https://patents.google.com/patent/WO2025208140A1/en = Xiaoyu Ma David Patterson. Google arXiv Jan [--] [----] https://t.co/lvUhAzkOqo "Tree-based Network Architecture for Accelerating Machine Learning Collective Operations" Nov [----] (Jan 2024) https://t.co/R5vQkMJY1j LLM with 3D-DRAM Chiplets (P:"
X Link 2026-01-17T07:23Z [----] followers, [---] engagements
"= "Critical Minerals on the U.S. Outer Continental Shelf: The Bureau of Ocean Energy Managements Role and Issues for Congress" Jan [--] [----] CRS EO [-----] Apr [----] Mineral Commodity Summaries [----] USGS https://x.com/ogawa_tter/status/1886958547482894345 https://www.whitehouse.gov/presidential-actions/2025/04/unleashing-americas-offshore-critical-minerals-and-resources/ https://www.congress.gov/crs-product/R48302 = Decision to Implement Export Control on Items Related to W Te Bi Mo & In Feb [--] [----] https://t.co/x4BRohlGEq USGS Mineral Commodity Summaries [----] Jan [--] https://t.co/9VORlON5Wy Ga&Ge"
X Link 2026-01-23T11:11Z [----] followers, [---] engagements
"Water Bottle Memory (2T Gain Cell Memory) Shuhan Liu Stanford 2nd MemoryDAX/DAM Winter WS Jan [--] [----] IEDM25 w/ TSMC NVIDIA Oxide Semicon for Memory Review Jan [--] = 3T GC-RAM https://x.com/ogawa_tter/status/2014664852078526734 https://x.com/ogawa_tter/status/2010608213436289284 https://x.com/ogawa_tter/status/1996939609570529777 https://dam.stanford.edu/assets/2026-workshop/dam-ww26-liu.pdf = "Design Optimization and Verification of Embedded Gain Cell RAMs" PhD Thesis [----] https://t.co/sfAOVXgrCM 3T GC-RAM in [--] nm CMOS [--] nm FDSOI [--] nm FinFET https://t.co/Gi0kaVQJqD RAAAM Memory"
X Link 2026-01-26T11:35Z [----] followers, [----] engagements
"= "A Device and Method for Reducing Memory Access Delay of 3D Stacked DRAM" NUDT (CN: Nov 2024) TSV & the Row buffers and Physical circuits form a high-speed data exchange bridge Stacked DRAM w/o TSV Huawei Appl (CN: Apr 2024) https://x.com/ogawa_tter/status/2004922428020183375 https://patents.google.com/patent/CN119938553A/en = "Semiconductor Packaging Structure and Electronic Device" Huawei Appl (CN: Apr 2024) https://t.co/iZuCutuRbj Stacked DRAM (400) without TSV Appls https://t.co/PrbrXhve4d DRAM (CN: Nov 2021) Stacked DRAM (CN: Dec 2021) Proprietary HBM: HiBL / HiZQ HUAWEI CONNECT 2025"
X Link 2026-01-31T12:28Z [----] followers, [---] engagements
"= "Thermal Management Challenges in 2.5D and 3D Chiplet Integration: A Review on ArchitectureCooling Co-Design" Eng Dec [--] [----] [--] pp [---] ref HB & HBM Jan [--] [----] HP & NVIDIA Microfluidic Cooling https://x.com/ogawa_tter/status/1971941691994300867 https://x.com/ogawa_tter/status/2017233675826675880 https://www.mdpi.com/2673-4117/6/12/373 = Trump administration cancels $45 million Oregon State grant for Microfluidics research (CorMic) May [--] [----] https://t.co/oQUQSssdKj OSU https://t.co/sxwqj4ME2i CHIPS & Science Act EDA Oct [----] https://t.co/7YerhcPjdu HP & NVIDIA ARPA-E COOLERCHIPS"
X Link 2026-02-02T14:39Z [----] followers, [---] engagements
"@1p_semicon @InoueFumihiro IEEE Hybrid Bonding Symposium Jan [--] [----] John H Lau (Unimicron) HB ()/ https://attend.ieee.org/hbs/page_id=779 https://x.com/ogawa_tter/status/2018165677656539340 = "From Interface Chemistry to Bond Wave: Unraveling SiCN/Cu Bonding Mechanisms" @InoueFumihiro YNU IEEE Hybrid Bonding Symp Jan [--] [----] (38:30) https://t.co/4PBaSUj5qQ https://t.co/glioDMnq5x https://t.co/9LvQUQQjhX Recent Advances & Trends Jan [--] https://t.co/XZhboN2mvn https://t.co/Z8BJVv6IVs https://attend.ieee.org/hbs/page_id=779 https://x.com/ogawa_tter/status/2018165677656539340 = "From Interface"
X Link 2026-02-03T03:25Z [----] followers, [----] engagements
"= [----] COOLERCHIPS Annual Review Project OMNICOOL Ali Heydari NVIDIA Nov [--] Ali Heydari Microfluidic Silicon Cold Plate HP & NVIDIA Nov [--] Trump Admin cancels $45M May [--] [----] https://x.com/ogawa_tter/status/2018333361459347831 https://arpa-e.energy.gov/sites/default/files/2026-01/Day2_03a_HPI_Cumbie.pdf https://x.com/ogawa_tter/status/1884921947131420948 https://arpa-e.energy.gov/sites/default/files/2026-01/Day1_04a_NVIDIA_Manaserh.pdf = "Thermal Management Challenges in 2.5D and 3D Chiplet Integration: A Review on ArchitectureCooling Co-Design" Eng Dec [--] [----] https://t.co/viXciJft0G [--] pp"
X Link 2026-02-03T10:58Z [----] followers, [---] engagements
"= "Package Architectures having Vertically Stacked Dies for High Capacity Memory" Intel (Dec 2023) High Capacity & High Bandwidth Memory while reducing the complexity of the package relative to conventional approaches J. Fryman = https://x.com/ogawa_tter/status/1704435232189882434 https://patents.google.com/patent/US20250201793A1/en = Intel. DARPA ERI Summit [----] Aug [--] Pat Gelsinger Opening https://t.co/pNO4dMbmXD https://t.co/LBK1e7iGdt Univ Shuttle https://t.co/HEdgpQs151 by Complex 3D Microsystems J. Fryman Intel Fellow https://t.co/HKPydLXJpo https://t.co/FWfmKIcxAR Sep 19"
X Link 2026-02-04T13:28Z [----] followers, [---] engagements
"@1p_semicon ( Advanced Memory Technology (AMT) DoE NNSA Intel AMT Jan [--] AMT (Sandia) https://x.com/ogawa_tter/status/1723159963995083206 = Projects for the Advanced Memory Technologies and Vendor R&D Product ASC Aug [--] [----] https://t.co/pRtfKJ1Jx5 R&D (SNL) Cerebras Oct [--] [----] https://t.co/3VNqEJY4JM Intel Dec [--] https://t.co/AdoTiDEWNK AMD Jan [--] [----] https://t.co/3kC5Vwg7W3 NVIDIA https://t.co/LrPEAPdKcL https://t.co/ovSx7e2BEp https://x.com/ogawa_tter/status/1723159963995083206 = Projects for the Advanced Memory Technologies and Vendor R&D Product ASC Aug [--] [----] https://t.co/pRtfKJ1Jx5 R&D"
X Link 2026-02-05T01:21Z [----] followers, [---] engagements
"= Positron AI Raises $230M Series B Feb [--] [----] Asimov HWed SoftMax RMSNorm RoPE SwiGLU Multiple on-chip ARMv9 Scale-Out Interconnect Credo OmniConnect Core AXI over 112G VSR = Weaver: LPDDR *) Credo & Positron AI On-Demand https://x.com/ogawa_tter/status/2012383715750711513 https://www.businesswire.com/news/home/20260204250472/en/Positron-AI-Raises-%24230-Million-Series-B-at-Over-%241-Billion-Valuation-to-Scale-Energy-Efficient-AI-Inference = IBM Buffered Server Memory POWER9 Power10: [----] GB/s OMI DDR4 Power11 Jul [----] https://t.co/cWKGugCfg6 [----] GB/s OMI DDR5 [--] TB [----] GB/s/Skt Credo"
X Link 2026-02-05T04:04Z [----] followers, [---] engagements
"Vertically Stacked Dies with a Cooling Microchannel Intel Appl (Sep 2023) = Thermal Mgmnt Challenges in 3D Review Dec [----] J. Fryman Intel ERI Summit [----] SAIMEMORY Feb [--] [----] https://x.com/ogawa_tter/status/2019249282877125075 https://x.com/ogawa_tter/status/1704435232189882434 https://x.com/ogawa_tter/status/2018333361459347831 https://patents.google.com/patent/US20250079263A1/en "SAIMEMORY"202623 https://t.co/O2FnHYEy8C [----] Eng https://t.co/YgWteRQQZV = https://t.co/B6ghH06aS2 Intel Feb [--] EE Times [--] AMT (NNSA/Sandia) Jan [--] Intel Appl (Dec 2023) https://t.co/hhekUv4KS8"
X Link 2026-02-05T11:07Z [----] followers, [----] engagements
"SAIMEMORY [------] [--] [--] NEXT S. Matsuoka Oct [----] = Fujitsu (Intel Appl) MONAKA-X+ZAM https://x.com/ogawa_tter/status/2019249282877125075 https://x.com/ogawa_tter/status/2018518028699464016 https://x.com/ogawa_tter/status/1983821426307297402 https://xtech.nikkei.com/atcl/nxt/column/18/00001/11457/P=2 https://txbiz.tv-tokyo.co.jp/txn/news_txn/post_334408 "SAIMEMORY"202623 https://t.co/O2FnHYEy8C [----] Eng https://t.co/YgWteRQQZV = https://t.co/B6ghH06aS2 Intel Feb [--] EE Times [--] AMT (NNSA/Sandia) Jan [--] Intel Appl (Dec 2023) https://t.co/hhekUv4KS8"
X Link 2026-02-06T12:20Z [----] followers, [----] engagements
"= Vertically Stacked Dies Intel Appls Voltage (Sep 2023) (Dec 2023) Power Delivery Cooling (Sep 2023) SAIMEMORY Feb [--] [----] High Capacity Memory (Dec 2023) https://x.com/ogawa_tter/status/2019367145742954949 https://patents.google.com/patent/US20250201768A1/en https://patents.google.com/patent/US20250201766A1/en https://patents.google.com/patent/US20250210587A1/en https://patents.google.com/patent/US20250079398A1/en Vertically Stacked Dies with a Cooling Microchannel Intel Appl (Sep 2023) https://t.co/qnCwLvlvbm = Thermal Mgmnt Challenges in 3D Review Dec [----] https://t.co/xUWfuMTxOM J. Fryman"
X Link 2026-02-06T14:40Z [----] followers, [---] engagements
"= Honda and Mythic: Joint Development of 100x Energy-Efficient Analog AI Chip for Next-Gen Vehicles Feb [--] [----] An automotive-grade AI SOC for deployment in Hondas next-gen SDVs by the late 2020s/early 2030s raised $125M Dec [--] https://mythic.ai/whats-new/mythic-to-challenge-ais-gpu-pantheon-with-100x-energy-advantage-and-oversubscribed-125m-raise/ https://mythic.ai/whats-new/honda-and-mythic-announce-joint-development-of-100x-energy-efficient-analog-ai-chip-for-next-generation-vehicles/"
X Link 2026-02-07T09:12Z [----] followers, [---] engagements
"= "Storage and Device" Huawei Appl (CN: Jun 2023) Stacked Storage Medium Wafer & Mmedium Controller Wafer Huawei Appls DRAM (CN: Nov 2021) "Wafer-on-Wafer Formed Memory and Logic" Micron (P: Aug 2021) https://x.com/ogawa_tter/status/1829500736302129505 https://x.com/ogawa_tter/status/1970494560678879673 https://patents.google.com/patent/US20260017196A1/en = Micron: Patent Applications (P: Aug 2021) "Wafer-on-Wafer Formed Memory and Logic" https://t.co/T2QK1beHb9 Logic devices can be AI accelerators can provide a greater bandwidth Testing https://t.co/9pNwTixzgN Signal Routing Input/Output"
X Link 2026-02-08T04:56Z [----] followers, [---] engagements
"= "Reviving Moore's Law with an EUV FEL" Nicholas Kelez CEO & CTO xLight Keynote EUV-FEL WS Jan [--] [----] Signs $150M LoI with the US DoC Dec [--] [----] K-Program: = https://x.com/ogawa_tter/status/2019582917706727722 https://x.com/ogawa_tter/status/1995759750395703519 https://conference-indico.kek.jp/event/350/contributions/7410/attachments/4960/8250/Presentation_Nicholas.pdf = Development of Next-Generation Semiconductor Process Technologies for EUV and BEUV under Japans K Program by JST Keynote 10th EUV-FEL WS Jan [--] [----] https://t.co/Qx6d1mkYQW [--] projects JPY 13.5B [--] yrs"
X Link 2026-02-09T02:48Z [----] followers, [---] engagements
"= Xian UniIC () Initiates IPO Preparations A Core Force in Domestic Memory Advances into the Capital Market Jan [--] [----] SeDRAM Nov [----] https://www.unisemicon.com/index.phpm=content&c=index&a=show&catid=24&id=181 https://x.com/ogawa_tter/status/1926825927574499677 http://en.unisemicon.com/product/qrsdram.html https://xinsanban.eastmoney.com/Article/NoticeContentid=AN202601061816086745 https://www.eet-china.com/mp/a465353.html = "A 3D Unified Analysis Method (3D-UAM) for Wafer-on-Wafer Stacked Near-Memory Structure" USTC & Xian UniIC Semicon IEEE Trans on VLSI System May [--] [----] (Early)"
X Link 2026-02-09T06:47Z [----] followers, [---] engagements
"= "The Vast Light Source Building the world's most powerful lasers " Y. Honda Accelerator Lab KEK Invited EUV-FEL WS Jan [--] [----] EUV-FEL WS 2024/3/2 Jan [--] xLight K-Program https://x.com/ogawa_tter/status/2019582917706727722 https://x.com/ogawa_tter/status/2020691317278535797 https://x.com/ogawa_tter/status/1801163094103589201 https://conference-indico.kek.jp/event/350/contributions/7404/attachments/4957/8241/Presentation_Honda.pdf = Development of Next-Generation Semiconductor Process Technologies for EUV and BEUV under Japans K Program by JST Keynote 10th EUV-FEL WS Jan [--] 2026"
X Link 2026-02-09T09:15Z [----] followers, [---] engagements
"= "Uni-STC: Unified Sparse Tensor Core" CUPB & NUDT HPCA [----] (PPTX) Sparse kernel: SpMV SpMSpV SpMM & SpGEMM Dataflow: Outer-product plus Segmented dot-product (A group of parallel vector mul. vector to update scalars)/cycle https://www.ssslab.cn/assets/slides/2026-lian-UniSTC.pptx https://www.ssslab.cn/assets/papers/2026-lian-UniSTC.pdf https://www.ssslab.cn/assets/slides/2026-lian-UniSTC.pptx https://www.ssslab.cn/assets/papers/2026-lian-UniSTC.pdf"
X Link 2026-02-10T03:55Z [----] followers, [---] engagements
"= "Characterizing Matrix Multiplication Units across General Parallel Patterns in Scientific Computing" CUPB & BSC PPoPP [----] NVIDIA A100/H200/B200 GEMV/GEMM/SpMV/SpGEMM FFT/Stencil/Reduction/Scan/BFS/PiC https://zenodo.org/records/17725527 https://www.ssslab.cn/assets/slides/2026-lu-Cubie.pdf https://dl.acm.org/doi/10.1145/3774934.3786456 https://zenodo.org/records/17725527 https://www.ssslab.cn/assets/slides/2026-lu-Cubie.pdf https://dl.acm.org/doi/10.1145/3774934.3786456"
X Link 2026-02-10T06:56Z [----] followers, [---] engagements
"= "High Bandwidth Destructive Read embedded Memory" Intel Granted: Jun [----] (Jun 2019) 2019-06-29: Application filed by Intel 2025-09-18 Assignment Owner name: SK HYNIX NAND PRODUCT SOLUTIONS CORP. (DBA SOLIDIGM) CALIFORNIA Effective date: [--------] 🤔 https://x.com/ogawa_tter/status/1599027773552885762 = "High Bandwidth Destructive Read embedded Memory" Intel Granted: Jun [----] (Jun 2019) https://t.co/blqOPqK3xx Low-area-overhead increases to the B/W of destructive read embedded memories (eDRAM) such as for caching https://t.co/lHIOnDWXWP DRAM NVIDIA https://t.co/kEl5eyV6Y0"
X Link 2026-02-10T07:32Z [----] followers, [---] engagements
"RT @suekichiii:"
X Link 2026-02-10T12:31Z [----] followers, [---] engagements
"RT @cqcqcqdx: First Public Demonstration of the Computer Mouse and Graphical User Interface [----] On December [--] [----] in what would late"
X Link 2026-02-10T13:19Z [----] followers, [---] engagements
"= "MXFP: A Metadata-Augmented Microscaling Data Format for Efficient Low-bit Quantization" SJTU & Computing Product Line Huawei ASPLOS [----] Mar [--] For Huawei Ascend 950PR (26 Q1)/950DT (26 Q4) FP8/MXFP8/HIF8: [--] PF MXFP4: [--] PF https://x.com/ogawa_tter/status/1968787152977674510 https://arxiv.org/abs/2601.19213v2 = HUAWEI CONNECT [----] Sep [--] https://t.co/6QXsLjPFU6 Ascend 950PR (Prefill & Recommend) [----] Q1 950DT (Decode & Training) [----] Q4 [--] PF (FP8)/2 PF (FP4) FP32/HF32/FP16/BF16/FP8/MXFP8/HiF8/MXFP4 Proprietary HBM NVIDIA Rubin CPX https://t.co/Hi2mWyQqeT Disaggregated"
X Link 2026-02-10T14:01Z [----] followers, [---] engagements
"RT @bcantrill: Our $200M Series C https://oxide.computer/blog/our-200m-series-c https://oxide.computer/blog/our-200m-series-c"
X Link 2026-02-10T14:50Z [----] followers, [--] engagements
"RT @InstLatX64: My current knowledge of #AMD #EPYC topologies: #Zen #Zen2 #Zen3 #Zen4 #Zen5 #Zen6 #Zen7"
X Link 2026-02-10T14:58Z [----] followers, [--] engagements
"RT @yuya0240:"
X Link 2026-02-10T14:59Z [----] followers, [---] engagements
"RT @aki_goodspeed: ✳"
X Link 2026-02-10T15:09Z [----] followers, [---] engagements
"RT @ZENRIN_official: 281"
X Link 2026-02-11T01:31Z [----] followers, [----] engagements
"RT @Mak0Nakamura: ()NIKE"
X Link 2026-02-11T04:19Z [----] followers, [---] engagements
"= "Full duplex DRAM for Tightly Coupled Compute Die and Memory Die" R. Osborne Intel (Aug 2022) HASWELL IEEE Micro Mar/Apr [----] eDRAM: Full-duplex CXL Full-duplex Google Appl https://x.com/ogawa_tter/status/2002744203726700723 https://x.com/ogawa_tter/status/2003433315152576960 https://pages.cs.wisc.edu/rajwar/papers/ieee_micro_haswell.pdf https://patents.google.com/patent/US20220392519A1/en = "Addressing Memory Contention in Modern DRAM-Based Systems" M. Patel Rutgers SAFARI Conf Dec [--] [----] https://t.co/Al5ItSQcAN https://t.co/bXiQx1A8Q4 PhD Thesis [----] https://t.co/BiauTLdXfy @onurmutlu"
X Link 2026-02-11T08:35Z [----] followers, [---] engagements
"RT @haskap1017:"
X Link 2026-02-11T09:33Z [----] followers, [----] engagements
"RT @S_Y_yama:"
X Link 2026-02-11T14:27Z [----] followers, [---] engagements
"RT @taku_photo47: 2"
X Link 2026-02-11T15:01Z [----] followers, [---] engagements
"RT @VRGSKtaQwz7ECBb:"
X Link 2026-02-11T15:39Z [----] followers, [---] engagements
"RT @FelixCLC_: This is my quarterly request to @GoogleAI to officially spec BF16 ideally via OCP. This is my quarterly request to @NVIDI"
X Link 2026-02-11T15:41Z [----] followers, [--] engagements
"RT @ASMLcompany: Our extreme ultraviolet (EUV) lithography systems print features just [--] nm wide orders of magnitude smaller than the le"
X Link 2026-02-11T18:44Z [----] followers, [---] engagements
"RT @tameiki: AI"
X Link 2026-02-12T01:41Z [----] followers, [--] engagements
"@yyasuda ( [----] [--] DRAMIntel FugakuNext ECC (RAS) Full duplex DRAM (2/5) https://x.com/ogawa_tter/status/2021503395916546317 = "Full duplex DRAM for Tightly Coupled Compute Die and Memory Die" R. Osborne Intel (Aug 2022) https://t.co/78SS6h6D0e HASWELL IEEE Micro Mar/Apr [----] https://t.co/FnUPZnMuXU eDRAM: Full-duplex CXL Full-duplex https://t.co/kbVDaE63ye https://t.co/viHtbBccqt Google Appl https://t.co/0LGhuw7JOS https://x.com/ogawa_tter/status/2021503395916546317 = "Full duplex DRAM for Tightly Coupled Compute Die and Memory Die" R. Osborne Intel (Aug 2022) https://t.co/78SS6h6D0e"
X Link 2026-02-12T01:52Z [----] followers, [--] engagements
"@yyasuda ( MONAKA-X (MSHR)NoC DoE ECP CPU+HBM (1st [--] 2) (3/5) https://x.com/ogawa_tter/status/1591674872253779968 = "HBM2/3 Evaluation on Many-Core CPUs" Jan [--] [----] https://t.co/HNLIP8PKaH Report Jun [--] [----] https://t.co/1HxW1Dg8it HPCG (XSBench) "higher NoC bandwidth and larger MSHR banks are critical for fully exploiting the performance benefits of HBM3" MSHR https://t.co/W00pkKyrsN https://t.co/Xm9icTpVXN https://x.com/ogawa_tter/status/1591674872253779968 = "HBM2/3 Evaluation on Many-Core CPUs" Jan [--] [----] https://t.co/HNLIP8PKaH Report Jun [--] [----] https://t.co/1HxW1Dg8it HPCG (XSBench)"
X Link 2026-02-12T01:53Z [----] followers, [--] engagements
"= NEDO20251218 [--] () SAIMEMORY202623 FugakuNEXT25 MONAKA-X & Intel ZAM (Full duplex DRAM) + NVIDIA GPU Stacked LLC (GC Memory) https://x.com/ogawa_tter/status/2015416350328656371 https://x.com/ogawa_tter/status/2021503395916546317 https://x.com/ogawa_tter/status/2020398753103093990 = "Three Dimensional Chip and Package Integration with Backside Metallization" NVIDIA (Sep 2024) https://t.co/IgRg0rTB4J https://t.co/ZeNTjatJiT COPA-GPU: Patent Mar [----] (P: Feb 2021) / ACM TACO Face-to-Face Dies Patents (Jan 2020) GC Memory IEDM25 https://t.co/DOugYZznRW https://t.co/yN3I3tz9RQ"
X Link 2026-02-12T08:40Z [----] followers, [---] engagements
"= Mueon Emerges From Stealth with $15.5M Seed to Redefine DCs for the AI Era Sep [--] [----] Cubelets Wilfred Gomes Co-founder & CEO Ponte Vecchio Vertically Stacked Dies Intel Appls https://x.com/ogawa_tter/status/2019367145742954949 https://x.com/ogawa_tter/status/2019783332700029334 https://x.com/ogawa_tter/status/1511475822057582593 https://www.intelcapital.com/mueon-emerges-from-stealth-with-15-5m-seed-to-redefine-data-centers-for-the-ai-era/ Vertically Stacked Dies with a Cooling Microchannel Intel Appl (Sep 2023) https://t.co/qnCwLvlvbm = Thermal Mgmnt Challenges in 3D Review Dec 2025"
X Link 2026-02-12T13:13Z [----] followers, [---] engagements
"@1p_semicon ( Intel "Package Architectures having Vertically Stacked Dies for High Capacity Memory" (202312) "Memory Layers Bonded to Logic Layers with Inclination" [-----] DRAM/SRAM ( https://x.com/ogawa_tter/status/2022156337065734249 = Vertically Stacked Dies Intel Appls "Memory Layers Bonded to Logic Layers with Inclination" (Sep 2023) https://t.co/rjOYJVFS9H (Dec 2023) High Capacity Memory https://t.co/NrDgERzT7l High Cap Mem for Power Delivery https://t.co/mMgmSFhWHP https://t.co/we3S9U15KE J. Fryman https://t.co/DALMNfWjz5 https://x.com/ogawa_tter/status/2022156337065734249 = Vertically"
X Link 2026-02-13T09:12Z [----] followers, [--] engagements
"@1p_semicon ( Intel [-----] "Full duplex DRAM for Tightly Coupled Compute Die and Memory Die" Randy B. Osborne HASWELL eDRAM (Full duplex DRAM) ZAM ( https://x.com/ogawa_tter/status/2022175488782086654 = Systems & Methods for HBM with Unidirectional Data Flow Google (P: Aug 2023) https://t.co/vt6AOFo1Ef Concurrent R and W Intel Appls Full duplex DRAM https://t.co/7twloembBI Vertically Stacked Dies https://t.co/yfXsOZf44i Split Write DRAM MEMSYS [----] https://t.co/qrALtS0MlL https://t.co/uBOuv33PFD https://x.com/ogawa_tter/status/2022175488782086654 = Systems & Methods for HBM with"
X Link 2026-02-13T09:14Z [----] followers, [--] engagements
"@1p_semicon ( [---] Mueon Co-founder & CEO W. Gomes Intel FellowPVC/Vertically Stacked DiesIntel Capital $15.5M Seed "Cubelets" Vertically Stacked Dies https://x.com/ogawa_tter/status/2021935819871899826 = Mueon Emerges From Stealth with $15.5M Seed to Redefine DCs for the AI Era Sep [--] [----] https://t.co/CpX6V2wK48 Cubelets Wilfred Gomes Co-founder & CEO Ponte Vecchio https://t.co/7K92LswVpn Vertically Stacked Dies Intel Appls https://t.co/mMgmSFhWHP https://t.co/we3S9U15KE https://t.co/jI4bznImdt https://x.com/ogawa_tter/status/2021935819871899826 = Mueon Emerges From Stealth with $15.5M Seed"
X Link 2026-02-13T09:16Z [----] followers, [--] engagements
"= [--------] SAIMEMORY202623 NEXT25 FugakuNEXT MONAKA-X w/ Intel ZAM & Full duplex DRAM +NVIDIA GPU Stacked LLC (GC Memory) https://x.com/ogawa_tter/status/2015416350328656371 https://x.com/ogawa_tter/status/2021503395916546317 https://x.com/ogawa_tter/status/2022156337065734249 https://x.com/ogawa_tter/status/2020398753103093990 = "Three Dimensional Chip and Package Integration with Backside Metallization" NVIDIA (Sep 2024) https://t.co/IgRg0rTB4J https://t.co/ZeNTjatJiT COPA-GPU: Patent Mar [----] (P: Feb 2021) / ACM TACO Face-to-Face Dies Patents (Jan 2020) GC Memory IEDM25"
X Link 2026-02-13T12:36Z [----] followers, [---] engagements
"= Neurophos: $110M Series A to Launch Exaflop-Scale Photonic AI Chips Jan [--] [----] Nov [----] $7.2M Seed Dec [--] [----] (P: Jan 2024) Sep [----] (P: Aug 2019) https://patents.google.com/patent/US11120332B2/en https://patents.google.com/patent/WO2025151899A1/en https://otc.duke.edu/news/neurophos-raises-7-2-million-round-to-create-super-fast-energy-efficient-metamaterial-based-optical-ai-chips-for-data-centers/ https://www.neurophos.com/whitepapers https://www.neurophos.com/110m-raise https://patents.google.com/patent/US11120332B2/en https://patents.google.com/patent/WO2025151899A1/en"
X Link 2026-01-26T08:04Z [----] followers, [---] engagements
"= "Recent Advances and Trends in Hybrid Bonding" John H Lau Unimicron Technology IEEE Hybrid Bonding Symposium Jan [--] [----] (29:21) State of the Art and Outlooks of CuCu HB J. Lau. IEEE TCPMT Oct [--] https://ieeexplore.ieee.org/abstract/document/11218187 https://attend.ieee.org/hbs/wp-content/uploads/sites/692/2026-1-1-Lau.pdf https://ieeetv.ieee.org/hbs/recent-advances-and-trends-in-hybrid-bonding https://ieeexplore.ieee.org/abstract/document/11218187 https://attend.ieee.org/hbs/wp-content/uploads/sites/692/2026-1-1-Lau.pdf"
X Link 2026-01-28T03:05Z [----] followers, [----] engagements
"= "Co-Packaged Optics - Heterogeneous Integration of Chiplets (Switch Photonic IC and Electronic IC)" John H Lau Unimicron J of Microelectronics & Elect Pkg Dec [--] [----] Co-Packaged Optics Book SemiAnalysis Jan [--] [----] https://x.com/ogawa_tter/status/2007777752204898431 https://imapsjmep.org/article/154275-co-packaged-optics-heterogeneous-integration-of-chiplets-switch-photonic-ic-and-electronic-ic = Co-Packaged Optics (CPO) Book: Scaling with Light for the Next Wave of Interconnect Jan [--] [----] @SemiAnalysis_ https://t.co/TT1tvvbkiZ 1: CPO TCO Analysis 2: CPO Intro & Implementation 3: Bringing"
X Link 2026-01-30T08:10Z [----] followers, [---] engagements
"= "Xscape Photonics and Tower Semiconductor Unveil the Industrys First Optically Pumped On-Chip Multi-Wavelength Laser Platform for AI Datacenter Fabrics" Aug [--] [----] OCP OSAI Sep [--] Appl (Oct 2023) https://x.com/ogawa_tter/status/1923670710142239136 https://x.com/ogawa_tter/status/1965784807604105442 https://towersemi.com/2025/08/25/08252025/ Next-Gen DRAM [----] Mar [----] Yole Gr https://t.co/t8co4SCF4B = System-Level WDM Switching for High Bandwidth & High-Capacity Memory Access Xscape Photonics (Oct 2023) https://t.co/a99PT5455b Oct [----] https://t.co/14KFeWqwew $44M Series A Oct [--] Google"
X Link 2025-11-04T14:59Z [----] followers, [----] engagements
"= IOWN APN2026113 NTT : IOWN [--------] [-----] https://x.com/ogawa_tter/status/1886914632964514119 https://journal.ntt.co.jp/article/37949 https://journal.ntt.co.jp/article/38187 https://www.soumu.go.jp/main_content/001050373.pdf = "TSMC NTT"202523 https://t.co/UMg1egcORN NTT20241023 https://t.co/21VDLp5fnZ https://t.co/8Ocjc8PQHg IOWN1213 https://t.co/Thl7cms3Kd Dec [--] / Jul [--] https://t.co/veeVrTY0s3 https://x.com/ogawa_tter/status/1886914632964514119 https://journal.ntt.co.jp/article/37949 https://journal.ntt.co.jp/article/38187 https://www.soumu.go.jp/main_content/001050373.pdf = "TSMC"
X Link 2026-02-02T09:17Z [----] followers, [---] engagements
"= GRAPH [---] June [----] BFS 1: Supercomputer Fugaku () [----] GTEPS: [-----] (K: 31302.4) Fujitsu A64FX Tofu-D 2: Sunway TaihuLight [----] [-------] 3: Summit (CPU-Only) [----] [------] SSSP 1: SuperMUC-NG 2: NERSC Cori [----] Haswell 3; Cray CS500 Xeon Phi http://graph500.org/ http://graph500.org/"
X Link 2020-06-22T08:54Z [----] followers, [--] engagements
"= Advanced Packaging AMD Case Study Tutorial Hot Chips [----] Aug [--] (25 pp) TSMC/Intel Chiplet G. Loh Aug [--] 3D Patent EFB Nov [--] https://x.com/ogawa_tter/status/1457819506458255365 https://x.com/ogawa_tter/status/1450522719536091138 https://x.com/ogawa_tter/status/1464220199020761093 https://x.com/ogawa_tter/status/1470641381458984965 https://hc33.hotchips.org/assets/program/tutorials/2021%20Hot%20Chips%20AMD%20Advanced%20Packaging%20Swaminathan%20Final%20%2020210820.pdf https://youtu.be/csAPSw11INUt=2193 = AMD Nov [--] [----] https://t.co/tltY8wXDoq AMD Instinct MI200: Elevated Fanout Bridge 2.5D"
X Link 2021-12-14T06:38Z [----] followers, [--] engagements
""Beyond [---] Gb/s High-speed Optical (and Copper) Interconnects" Samtec Invited TWEPP Sep [--] [----] PPTX (35 MB) https://indico.cern.ch/event/608587/contributions/2697522/attachments/1523073/2380259/TWEPP_2017_Samtec_Presentation_03.pptx https://indico.cern.ch/event/608587/contributions/2697522/attachments/1523073/2380259/TWEPP_2017_Samtec_Presentation_03.pptx"
X Link 2017-09-16T14:23Z [----] followers, [--] engagements
"= "Integrating NVIDIA Deep Learning Accelerator (NVDLA) with RISC-V SoC on FireSim" 2nd WS on Energy Efficient Machine Learning and Cognitive Computing for Embedded Applications (EMC2) Fe [--] [----] FireSim https://fires.im/ https://github.com/CSL-KU/firesim-nvdla https://docs.wixstatic.com/ugd/42e7ad_480c39fe4d4b4d7e878c8ded85041ad5.pdf https://fires.im/ https://github.com/CSL-KU/firesim-nvdla https://docs.wixstatic.com/ugd/42e7ad_480c39fe4d4b4d7e878c8ded85041ad5.pdf"
X Link 2019-02-17T21:30Z [----] followers, [--] engagements
"= "Shasta Hardware Technical Workshop" Cray CUG [----] May [--] [----] PDF [--] pp Mountain / River (19" Rack) Slingshot NIC: "Mellanox" Rosetta: Broadcom IP Exascale Day 2019/1018 Steve Scott Sep [----] = https://x.com/ogawa_tter/status/1171793483889119232 https://x.com/ogawa_tter/status/1186759552114388992 https://cug.org/proceedings/cug2019_proceedings/includes/files/inv112s1-file1.pdf = "The Cray Shasta Architecture" Steve Scott Cray HPC User Forum Sep [--] [----] https://t.co/iI95MNXz8c Slingshot: Gnodal Patents (Hot Interconnects 2019) https://t.co/yvJX3cTdmk NERSC Perlmutter https://t.co/sb3VqGAbdx"
X Link 2019-11-17T11:55Z [----] followers, [--] engagements
"= "GPUDirect Storage: Transfer Data Directly to GPU Memory Alleviating IO Bottlenecks" NVIDIA SC19 Nov [--] [----] Video Slides RAPIDS Sep [----] GPUDirect Storage Blog Aug [----] https://x.com/ogawa_tter/status/1159401097200459776 https://x.com/ogawa_tter/status/1176605978227142656 https://on-demand.gputechconf.com/supercomputing/2019/pdf/sc1922-gpudirect-storage-transfer-data-directly-to-gpu-memory-alleviating-io-bottlenecks.pdf https://on-demand.gputechconf.com/supercomputing/2019/video/sc1922-gpudirect-storage-transfer-data-directly-to-gpu-memory-alleviating-io-bottlenecks/ = "GPUDirect Storage:"
X Link 2019-12-13T02:17Z [----] followers, [--] engagements
"= ""JAMSTEC201012 PDF via @keaton_n18 Computers in Physics 1992/88 Supercomputing in Japan Oyanagi [----] [-----] https://x.com/ogawa_tter/status/1220575846206857216 https://x.com/ogawa_tter/status/1219333793674170368 https://x.com/ogawa_tter/status/1215750543731068928 http://www.jamstec.go.jp/es/jp/publication/pdf/Development_ES.pdf https://x.com/ogawa_tter/status/1220575846206857216 https://x.com/ogawa_tter/status/1219333793674170368 https://x.com/ogawa_tter/status/1215750543731068928 http://www.jamstec.go.jp/es/jp/publication/pdf/Development_ES.pdf"
X Link 2020-06-24T06:11Z [----] followers, [--] engagements
"= "PIUMA: Programmable Integrated Unified Memory Architecture" Intel arXiv Oct [--] [----] A graph analysis oriented architecture Custom RISC ISA Hierarchical topology (HyperX) PIUMA IEEE HPEC [----] PUMA (DARPA HIVE) FOSDEM [----] https://x.com/ogawa_tter/status/1308870123973943296 https://arxiv.org/abs/2010.06277 = "Hash Table Scalability on Intel PIUMA (Programmable and Integrated Unied Memory Architecture)" Intel IEEE HPEC [----] Sep [--] [----] Efficient Sparse Matrix-Vector Multiplication on Intel PIUMA IEEE HPEC Sep [--] https://t.co/6tXhE80qbz PUMA (DARPA HIVE) Intel FOSDEM 2020"
X Link 2020-10-22T02:06Z [----] followers, [--] engagements
"= @LightOnIO Appliance a Photonic Co-processor sets a new pathway for Transformative AI / HPC computing Mar [--] [----] The First Commercial Photonic Computing in your own DC [--] TOPS/W https://lighton.ai/publications/ https://lighton.ai/wp-content/uploads/2021/03/LightOn-Appliance-spec-sheet.pdf https://lighton.ai/lighton-appliance/ https://lighton.ai/wp-content/uploads/2021/03/LightOn-Appliance-press-release.pdf https://lighton.ai/publications/ https://lighton.ai/wp-content/uploads/2021/03/LightOn-Appliance-spec-sheet.pdf https://lighton.ai/lighton-appliance/"
X Link 2021-03-08T14:27Z [----] followers, [--] engagements
"= AArch64-ExploreExploration of Apple CPUs @handleym99 Vol 1: CPU Core (205 pp) Vol 2: Load Store Unit Cache & Memory (215 pp) Vol 3: Patent Exploration (116 pp) My Apple patent tweets https://twilog.org/ogawa_tter/searchword=Apple%20patent&ao=a https://github.com/name99-org/AArch64-Explore https://x.com/handleym99/status/1511475655447232512 OK folks the big update is here Everything can be found at https://t.co/HDozWzQRHX I know nothing about Git or Github and have little interest in learning so I likely screwed up something but what you should find here is 1/"
X Link 2022-04-06T04:14Z [----] followers, [--] engagements
"= Intel & SoftBank Subsidiary SAIMEMORY Collaborate to Advance Next-Generation Memory for AI Feb [--] [----] Z-Angle Memory Jan [--] AMT Dec [----] Appl (Dec 2023) https://x.com/ogawa_tter/status/2019040243077570752 https://x.com/ogawa_tter/status/1603609370859737089 https://www.sandia.gov/research/news/from-early-rd-to-mature-technology/ https://eetimes.itmedia.co.jp/ee/articles/2602/03/news099.html https://community.intel.com/t5/Blogs/Intel/Policy-Intel/Intel-and-SoftBank-Subsidiary-SAIMEMORY-Collaborate-to-Advance/post/1735559 = "Package Architectures having Vertically Stacked Dies for High Capacity"
X Link 2026-02-04T13:31Z [----] followers, [----] engagements
""SAIMEMORY"202623 [----] Eng = Intel Feb [--] EE Times [--] AMT (NNSA/Sandia) Jan [--] Intel Appl (Dec 2023) https://x.com/ogawa_tter/status/2019041233503658372 https://www.saimemory.co.jp/en/news/press/20260203_01/ https://www.saimemory.co.jp/news/press/20260203_01/ = Intel & SoftBank Subsidiary SAIMEMORY Collaborate to Advance Next-Generation Memory for AI Feb [--] [----] https://t.co/uIHQPTh4bb https://t.co/EFa0mj0mL7 Z-Angle Memory Jan [--] https://t.co/4HF7hLvQcm AMT Dec [----] https://t.co/ch6fI6A3Vk Appl (Dec 2023) https://t.co/NrDgERzT7l https://t.co/j3U50UkvW0"
X Link 2026-02-05T03:18Z [----] followers, [----] engagements
"= Vertically Stacked Dies Intel Appls Stephen Morein CTO SAIMEMORY (Aug [----] --) w/ Solder Interconnects (Aug 2023) (Jun 2022) C Bricks Planarized Edges = https://patents.google.com/patent/US20230420409A1/en https://patents.google.com/patent/US20230420411A1/en https://patents.google.com/patent/US20240006375A1/en https://patents.google.com/patent/US20250062278A1/en https://x.com/ogawa_tter/status/2022156337065734249 = Vertically Stacked Dies Intel Appls "Memory Layers Bonded to Logic Layers with Inclination" (Sep 2023) https://t.co/rjOYJVFS9H (Dec 2023) High Capacity Memory"
X Link 2026-02-14T07:50Z [----] followers, [---] engagements
"= "Double Coil Inductive Scheme and Modulation for Z-Axis Memory (ZAM) Links" Intel Appl (Jul 2024) Stephen Morein CTO SAIMEMORY (Aug [----] --) SAIMEMORY Feb [--] [----] ZAM (Z-Angle Memory) AMT NNSA Jan [--] https://x.com/ogawa_tter/status/2019249282877125075 https://x.com/ogawa_tter/status/2022579290274144512 https://patents.google.com/patent/US20260005164A1/en "SAIMEMORY"202623 https://t.co/O2FnHYEy8C [----] Eng https://t.co/YgWteRQQZV = https://t.co/B6ghH06aS2 Intel Feb [--] EE Times [--] AMT (NNSA/Sandia) Jan [--] Intel Appl (Dec 2023) https://t.co/hhekUv4KS8"
X Link 2026-02-14T07:51Z [----] followers, [---] engagements
"Brain-Inspired Intelligent SoCs and Applications Hoi-Jun Yoo KAIST Keynote MCSoC-16 PDF http://ssl.kaist.ac.kr/2007/ http://mcsoc-forum.org/2016/wp-content/uploads/2015/07/keynote-Yoo.pdf http://ssl.kaist.ac.kr/2007/ http://mcsoc-forum.org/2016/wp-content/uploads/2015/07/keynote-Yoo.pdf"
X Link 2017-02-01T03:24Z [----] followers, [--] engagements
"= "Stream-Dataflow Acceleration" ISCA [----] Slides [---] pages Paper RISC-V based simulator https://x.com/ogawa_tter/status/883310226958450688 http://research.cs.wisc.edu/vertical/talks/2017/isca17-stream-dataflow-talk = "Stream-Dataflow Acceleration" Tony Nowatzki et al. ISCA [----] PDF https://t.co/LP3gHqOru3 Softbrain Cycle-level RISC-V based simulator https://t.co/H7LCvR923y https://x.com/ogawa_tter/status/883310226958450688 http://research.cs.wisc.edu/vertical/talks/2017/isca17-stream-dataflow-talk = "Stream-Dataflow Acceleration" Tony Nowatzki et al. ISCA [----] PDF https://t.co/LP3gHqOru3"
X Link 2017-08-08T11:47Z [----] followers, [--] engagements
"= "A Deep Dive Into NEC's Aurora Vector Engine" Nov [--] [----] MPI operations to be directory executed between the VEs w/o copying VE can access Xeon DDR4 via DMA 6x HBM2 1.2TB/s VecLen: [---] x 64bit 2.45TF @ 1.6GHz STREAM/node price HPL/node price https://www.nextplatform.com/2017/11/22/deep-dive-necs-aurora-vector-engine/ https://www.nextplatform.com/2017/11/22/deep-dive-necs-aurora-vector-engine/"
X Link 2017-11-22T23:56Z [----] followers, [--] engagements
"= "A case for scoped persist barriers in GPU" AMD Research GPGPU [----] Feb [--] [----] HW support for persist barriers The only additional HW is Non-Volatile-Write FIFOs AMD Radeon Pro SSG Datasheet HBCC http://pro.radeon.com/_downloads/radeon-pro-ssg-datasheet-27.7.17.pdf https://pro.radeon.com/en/product/pro-series/radeon-pro-ssg/ http://drona.csa.iisc.ac.in/arkapravab/papers/gpgpu18_scoped_persistent_barriers.pdf http://pro.radeon.com/_downloads/radeon-pro-ssg-datasheet-27.7.17.pdf https://pro.radeon.com/en/product/pro-series/radeon-pro-ssg/"
X Link 2018-03-06T11:30Z [----] followers, [--] engagements
""Tesla's Autopilot Chief Keller Steps Down After Two Years" Bloomberg Apr [--] [----] 12:17 JST Updated 13:18 JST "AMD CPU hero Jim Keller leaving Tesla to join Intel" TweakTown Apr [--] [----] JST "According to my industry sources" https://www.tweaktown.com/news/61675/amd-cpu-hero-jim-keller-leaving-tesla-join-intel/index.html https://www.bloomberg.com/news/articles/2018-04-26/tesla-says-autopilot-vice-president-keller-is-leaving-carmaker https://www.tweaktown.com/news/61675/amd-cpu-hero-jim-keller-leaving-tesla-join-intel/index.html"
X Link 2018-04-26T06:04Z [----] followers, [--] engagements
"= Qualcomm Acquires Ventana Micro Systems Deepening #RISCV CPU Expertise Dec [--] [----] "Ventana team will complement Qualcomms ongoing RISC-V and custom Oryon CPU development further advancing technology leadership in the AI era across all businesses." https://www.qualcomm.com/news/releases/2025/12/qualcomm-acquires-ventana-micro-systems--deepening-risc-v-cpu-ex https://www.qualcomm.com/news/releases/2025/12/qualcomm-acquires-ventana-micro-systems--deepening-risc-v-cpu-ex"
X Link 2025-12-11T02:18Z [----] followers, [---] engagements
"= ""NRIJST2025822 20258JICA (314) HPC https://x.com/ogawa_tter/status/1993869856249418069 https://x.com/ogawa_tter/status/1986425912309784825 https://spap.jst.go.jp/event/pdf/study045.pdf https://www.youtube.com/watchv=2lHaOqMcnnY = "India has satisfied its supercomputing needs but not its ambitions" Nov [--] [----] https://t.co/ALa8rlROEG [--] machines: [--] PF + [--] PF [----] National Supercomputing Mission RSDs [----] https://t.co/3zQC5ZMvtk Rudra Server AUM 5nm [--] cores: 2x 24x (2x Neoverse-V2) DDR5 HBM3 https://x.com/ogawa_tter/status/1993869856249418069"
X Link 2025-12-11T05:20Z [----] followers, [---] engagements
"= Memory: Almost The Only Thing That Matters A Revolution in Memory Arch for the DC Marvell Hot Chips [----] Custom SRAM Jun [--] Custom HBM Dec [----] [--] Gbps/wire BiDI D2D Aug [--] https://www.marvell.com/company/newsroom/marvell-unveils-64gbps-bidirectional-die-to-die-interface-ip-2nm-xpu.html https://x.com/ogawa_tter/status/1935576279216197659 https://hc2025.hotchips.org/assets/program/conference/day2/40_Marvell_Kuemerle_final.pdf https://www.youtube.com/watchv=p97UkNh2oPc&t=38s = Marvell Custom AI Investor Event Jun [--] [----] (2:32:41) https://t.co/F2VESxFs4t (124 pp) https://t.co/qPrcAtqXP0"
X Link 2025-12-11T14:24Z [----] followers, [---] engagements
"= "Accelerating AI with UALink: Open Memory Fabrics for Scalable Compute" R. Hyatt UnifabriX Dec [--] [----] Alibaba: ALink AMD Helios & HPE: UALoE UnifabriX https://x.com/ogawa_tter/status/1967956344792813837 https://x.com/ogawa_tter/status/1996037944759746638 https://x.com/ogawa_tter/status/1986634590325514505 https://ualinkconsortium.org/wp-content/uploads/2025/12/UnifabriX_UALink-Webminar_20251210_PDF.pdf https://www.youtube.com/watchv=N_8eVUd0utA = UnifabriX https://t.co/k5BiHYCMw0 Out of Stealth Mode Nov [--] [----] https://t.co/PP26HbEWrJ Lord of the Memory Fabrics ISW [----] Sep 10"
X Link 2025-12-12T05:44Z [----] followers, [---] engagements
"= United States and Japan Will Sign Pax Silica Preamble Dec [--] [----] Pax Silica Initiative: Fact Sheet Pax Silica Summit & Sony Samsung SK Hynix ASML etc https://www.state.gov/pax-silica/ https://www.state.gov/releases/office-of-the-spokesperson/2025/12/pax-silica-initiative/ https://x.com/UnderSecE/status/1999315601290822037 https://www.state.gov/releases/office-of-the-spokesperson/2025/12/united-states-and-japan-will-sign-pax-silica-preamble/ Honored to spearhead todays landmark summit in collaboration with @DeputySecState @DavidSacks47 Ethan Klein alongside other colleagues across the"
X Link 2025-12-12T11:05Z [----] followers, [---] engagements
"= "NeuronMM: High-Performance Matrix Multiplication for LLM Inference on AWS Trainium" arXiv Nov [--] (Oct 29) PSUM w/ near-memory accumulation to enables read-accumulate-write ops Trainium3 NeuronCore-v4 https://awsdocs-neuron.readthedocs-hosted.com/en/latest/about-neuron/arch/neuron-hardware/neuron-core-v4.html https://x.com/ogawa_tter/status/1997870805313876448 https://arxiv.org/abs/2510.25977 = AWS Trn3 UltraServers: Power next-gen enterprise AI performance AWS re:Invent [----] https://t.co/sqC5x5Ty0a HWed Microslcaling (MXFP8) & Softmax func Trainium4 NVIDIA Dec [--] https://t.co/k3M2NMSfn9"
X Link 2025-12-13T03:19Z [----] followers, [---] engagements
"= "Made in China 2025: Evaluating China's Performance" Nov [--] [----] U.S.-China Economic & Security Review Commission Scorecard Next-Gen Information Technology: Semiconductors . Made in China [----] [----] May [----] https://x.com/ogawa_tter/status/1501989064651640842 https://www.uscc.gov/research/made-china-2025-evaluating-chinas-performance = Translation: Notice of the State Council on the Publication of Made in China [----] CSET Mar [--] [----] https://t.co/PGfrToSPNh One of Chinas most important industrial policies [----] [------] https://t.co/wPwuZsBxFF https://t.co/u32cUVdwX2 https://t.co/yldNiZN6sL"
X Link 2025-12-13T07:42Z [----] followers, [---] engagements
"= "Magnetics Integrations for 2.5D and 3D Packaging" Tyndall Power Magnetics @ High Frequency [----] WS Mar [--] The necessity of 2.5D/3D integration of magnetics Classification of 3D integrated magnetics 2.5D HI of magnetics by micro-transfer technology https://psma.com/sites/default/files/uploads/MagneticsWorkshopPresentations/Magnetics_Integrations_for_2.5D_and_3D_Packaging__Sai_Tyndall_S.pdf https://psma.com/sites/default/files/uploads/MagneticsWorkshopPresentations/Magnetics_Integrations_for_2.5D_and_3D_Packaging__Sai_Tyndall_S.pdf"
X Link 2025-12-13T12:17Z [----] followers, [---] engagements
"= "Reimaging and Transforming Package Assembly and Test Manufacturing in the AI Era" SATAS () VP Foundry Tech Manufacturing Intel HIR Session at SEMICON West [----] Oct [--] NEDO [------] https://x.com/ogawa_tter/status/1854410029757329749 https://eps.ieee.org/images/files/HIR_Presentations/SEMICON_W_2025/Oct_8_Jeffrey_Pattinato.pdf https://x.com/ogawa_tter/status/1854410029757329749 https://eps.ieee.org/images/files/HIR_Presentations/SEMICON_W_2025/Oct_8_Jeffrey_Pattinato.pdf"
X Link 2025-12-14T04:02Z [----] followers, [--] engagements
"= "Feeding the Beast: The Critical Challenges of Powering Next-Gen AI accelerators" Jawad Haj-Yahya Rivos 1st @SAFARI_ETH_CMU Conf [----] Dec [--] @onurmutlu DC [---] VDC Vertical PD https://x.com/ogawa_tter/status/1973657582549213514 https://x.com/ogawa_tter/status/1981369625276551328 https://events.safari.ethz.ch/SafariConference25/lib/exe/fetch.phpmedia=jawad_future_of_pdn_2025.pptx https://www.youtube.com/live/QtS7v8aVGLQt=11530s = "Hybrid Switched-Capacitor Circuits and Magnetics Co-Design for Vertical Power Delivery" PhD Thesis [----] https://t.co/sHbynIDNNU Mini-LEGO Nov 2023"
X Link 2025-12-16T12:32Z [----] followers, [---] engagements
"= Federal gov terminates $285M contract for SMART USA CHIPS manufacturing institute Dec [--] [----] 'termination for convenience' Executive Director ORAU Webinar Oct [--] Feb [--] https://x.com/ogawa_tter/status/1963953570593771548 https://www.orau.org/partnerships/documents/webinar/digital-manufacturing-smart.pdf https://www.smartusatwins.org/ https://www.wral.com/news/local/federal-terminates-285m-contract-durham-chips-institute-december-2025/ = Digital Twin in Semicon How AI/ML Advances HI Keynote HIR Sessions ECTC [----] https://t.co/TAVObdxKbY SMART USA HIR Feb [--] https://t.co/gj8L6UeEHr Jan 3"
X Link 2025-12-17T05:04Z [----] followers, [----] engagements
"= "Chip Scan: 3D X-ray imaging of CMOS integrated circuits" A.F.J. Levi USC Keynote Microelectronics Reliability and Qualification WS (MRQW) Feb [--] [----] 4nm Nature Jul [--] [----] 22nm Apr [--] [----] https://x.com/ogawa_tter/status/1818874067765936260 https://www.nature.com/articles/s41586-024-07615-6.epdfsharing_token=wRppPTjk7NeAyZocQG2NkdRgN0jAjWel9jnR3ZoTv0MFXlcEC43QR4AJLMRV3QQH27tVRSk9jZqg2AuLkhsTaedXagMSnk7I3F3m2gJfmEJMcrxLcjnmlxxcZB5kOgr51a6rPpBo-UJ3x2dBdbogakoNROieAnvF1oQj1wDCpsk%3D"
X Link 2025-12-17T13:37Z [----] followers, [---] engagements
"= "Solving The Problems of HBM-on-Logic" Dec [--] [----] @gavbon86 (Guest) & @IanCutress imec at IEDM [----] Future AI Accelerators might Need to be Slower to be Faster HIFOM (Hynix Integrated Fan-out Memory) Logic on stacked DRAM https://x.com/ogawa_tter/status/1726944199168430108 https://morethanmoore.substack.com/p/solving-the-problems-of-hbm-on-logic = "Fan-out Wafer Level Package for Memory Applications" SK hynix Chip Scale Review Jan-Feb [----] https://t.co/norLW2GJYu ECTC [----] https://t.co/MpzVX2ZUD9 Patent May [----] (KR: Aug 2018) https://t.co/L5dfZew0vQ SK hynix Patent Apr [----] (KR: Aug 2020)"
X Link 2025-12-18T04:02Z [----] followers, [---] engagements
"= Loongson performance in benchmarks Tramplin Electronics and NEIMARK: LoongArch Architecture Meetup Dec [--] [----] Tramplin Electronics unveiled the Russian Irtysh C616 server processor Nov [--] based on LoongArch https://x.com/ogawa_tter/status/1995737047512793284 https://vk.com/neimark_itw=wall-211802468_1420 = Tramplin Electronics unveiled the Russian Irtysh C616 server processor Nov [--] [----] https://t.co/E5u7wxHW5G based on LoongArch architecture C616: [--] cores [---] GHz [-----] GF 632: . 664: [--] cores [---] GHz [----] GF https://t.co/wXQY20oIDx https://t.co/91L8nl7qsX https://t.co/oAjg6pHY0L"
X Link 2025-12-18T07:36Z [----] followers, [---] engagements
"= SCREEN Hokdings IR Day for FY2026/03 Dec [--] [----] ATCA: $75M Strategic Partnership w/ NY Creates IBM Sep [--] Nikon Oct [--] [-----] IR Day1216 https://www.screen.co.jp/download_file/view/88a2eb9f-6bcc-4ff6-acfb-7bea28f88252/449 https://www.screen.co.jp/en/news/NR251031E https://www.screen.co.jp/spe/en/information/spe250924 https://ny-creates.org/governor-hochul-announces-75-million-strategic-partnership-between-ny-creates-and-screen-to-strengthen-u-s-and-japan-semiconductor-rd-collaboration/ https://www.screen.co.jp/download_file/view/c840ea6a-82e5-4a4a-a930-77e0a8dc8d3f/1210"
X Link 2025-12-18T12:46Z [----] followers, [----] engagements
"= "Semiconductor challenges in ramping AI at the edge" PR Chidi Chidambaram SVP Engineering & Qualcomm Fellow Qualcomm Technologies HIR Session at SEMICON West [----] Oct [--] Device 3D + System 3D Hetero. integration STCO: New Language Needed https://eps.ieee.org/images/files/HIR_Presentations/SEMICON_W_2025/Oct_8_PR_Chidi_Chidambaram.pdf https://eps.ieee.org/images/files/HIR_Presentations/SEMICON_W_2025/Oct_8_PR_Chidi_Chidambaram.pdf"
X Link 2025-12-19T03:35Z [----] followers, [---] engagements
"= "Survey of storage systems in high performance computing" NUDT China Computer Federation (CCF) Transactions on HPC Dec [--] [----] Systematically reviews covering four aspects: Storage architecture Hardware Software and Networking [--] pp [--] () ref https://link.springer.com/article/10.1007/s42514-025-00268-5 https://link.springer.com/article/10.1007/s42514-025-00268-5"
X Link 2025-12-19T07:21Z [----] followers, [---] engagements
""Photonic Connectivity for Al Systems" K. Bergman. Columbia EFCL Int WS on Sustainable AI Nov [--] [----] 3D Photo OXC Jul [--] Beyond CLOS Huawei = HBM Columbia HotI [----] https://x.com/ogawa_tter/status/2001896001653932197 https://x.com/ogawa_tter/status/1985550170873122966 https://x.com/ogawa_tter/status/1904371742581408209 https://video.ethz.ch/events/2025/efcl/v/Jp2fo7TbBhw = "Solving The Problems of HBM-on-Logic" Dec [--] [----] https://t.co/QET8P02fUb HBM (Memory) with an optical connection Celestial AI https://t.co/FhgK8qU2Jb IEEE CAL https://t.co/HZ5uxuLmx3 Google Appls Columbia HotI 2025"
X Link 2025-12-19T13:23Z [----] followers, [---] engagements
"= ACTINA: Adapting Circuit-Switching Techniques for AI Networking Arch Columbia & NVIDIA SC25 Flexible Photo Accelerated High Performance Compute System Z. Wu PhD Thesis [----] = NVIDIA K. Bergman Nov [--] https://x.com/ogawa_tter/status/2002006934044680479 https://academiccommons.columbia.edu/doi/10.7916/m381-v267 https://dl.acm.org/doi/full/10.1145/3712285.3759842 "Photonic Connectivity for Al Systems" K. Bergman. Columbia EFCL Int WS on Sustainable AI Nov [--] [----] https://t.co/l9ifnh9xfg 3D Photo https://t.co/Zc7KbYUv9g OXC Jul [--] https://t.co/mJ6CjRoi7r Beyond CLOS Huawei = HBM"
X Link 2025-12-20T08:39Z [----] followers, [---] engagements
"= "Modularity at the Package Edge: Composable Interconnect for [---] Gb/s/Lane AI Systems" Bijan Nowroozi Head of Ecosystem Development Lightmatter OCP HPC Sub-Proj Dec [--] [----] Next-Gen OAM 400Gbs/lane fabric at the system level https://drive.google.com/file/d/1eQbxpCDgtlM1lH9JCyxUQ2cLPFc0b_fj/view https://www.youtube.com/watchv=CaIyMse50Jo https://drive.google.com/file/d/1eQbxpCDgtlM1lH9JCyxUQ2cLPFc0b_fj/view https://www.youtube.com/watchv=CaIyMse50Jo"
X Link 2025-12-21T10:41Z [----] followers, [---] engagements
"= NVIDIA CUDA Tile using cuTile and CUDA Tile IR to write portable high-performance GPU Kernels @JokerEph & @roeschinc GPU Mode Dec [--] [----] cuTile Python Dec [--] May [--] GTC [----] https://x.com/ogawa_tter/status/1925023993679290758 https://developer.nvidia.com/blog/simplify-gpu-programming-with-nvidia-cuda-tile-in-python https://developer.nvidia.com/cuda/tile https://www.youtube.com/watchv=_b4I4rKpsGA = "The Essence of CUDA C++ : Past Present and Future" V. Grover Sr. Distinguished Engineer NVIDIA Keynote AI in the Cloud IAP May [--] [----] https://t.co/cFw7Ca6f8h Mar [----] https://t.co/TcartJ9JCu CUDA"
X Link 2025-12-22T05:16Z [----] followers, [---] engagements
"= "Split Write DRAM: Reducing DRAM Access Latency by Mitigating Read-Write Interference" K Chitra (IIT Guwahati) Minesh Patel (Rutgers) MEMSYS [----] enable timely write draining while minimizing read disruption M. Patel Dec [--] https://x.com/ogawa_tter/status/2002744203726700723 https://www.memsys.io/wp-content/uploads/ninja-forms/5/Split_Write_DRAM.pdf = "Addressing Memory Contention in Modern DRAM-Based Systems" M. Patel Rutgers SAFARI Conf Dec [--] [----] https://t.co/Al5ItSQcAN https://t.co/bXiQx1A8Q4 PhD Thesis [----] https://t.co/BiauTLdXfy @onurmutlu R/W in parallel POWER9=Power10 Google"
X Link 2025-12-22T06:49Z [----] followers, [---] engagements
"= "CXLAimPod: CXL Memory is all you need in AI era" arXiv Aug [--] [----] Benefits of CXL's Full-duplex Yiwei Yang UCSC Split Write DRAM (Half) Full-duplex POWER9 Power10: Open Memory Interface Google Appl https://x.com/ogawa_tter/status/2002744203726700723 https://scholar.google.com/citationsuser=sDUT22MAAAAJ&view_op=list_works&sortby=pubdate https://arxiv.org/abs/2508.15980 = "Addressing Memory Contention in Modern DRAM-Based Systems" M. Patel Rutgers SAFARI Conf Dec [--] [----] https://t.co/Al5ItSQcAN https://t.co/bXiQx1A8Q4 PhD Thesis [----] https://t.co/BiauTLdXfy @onurmutlu R/W in parallel"
X Link 2025-12-23T11:51Z [----] followers, [---] engagements
"= "Memory Access Scheduling" ISCA@50 25-Year Retrospective Jun [----] ISCA [----] S. Rixner PhD Thesis [----] William J. Dally Stream Processors Mar [----] SPI https://en.wikipedia.org/wiki/Stream_Processors_Inc https://x.com/ogawa_tter/status/1987143958276059280 https://dspace.mit.edu/handle/1721.1/86579 https://dl.acm.org/doi/10.1145/339647.339668 https://sites.coecis.cornell.edu/isca50retrospective/files/2023/06/Rixner_2000_Memory.pdf = William James Dally [----] Benjamin Franklin Medal in Computer & Cognitive Science https://t.co/OtFWa4CqZP https://t.co/WucQGgalko Merrimac SC03 Brook for GPUs"
X Link 2025-12-23T14:32Z [----] followers, [---] engagements
"= "XEM: Tensor accelerator for AB21 supercomputing artificial intelligence processor" ETRI Journal Oct [--] [----] AB21: TSMC 12-nm. [-----] x [-----] m [-----] W XEM 4x4@FP64 / 8x8@FP32 4.58TF / 18.32TF [----] MHz AB21 MUG [----] https://x.com/ogawa_tter/status/1963190097635061895 https://onlinelibrary.wiley.com/doi/10.4218/etrij.2024-0141 = Supreme-K: A Korean Initiative for Pioneering the AB21 HPC Accelerator & Compute Node ETRI MUG [----] Aug [--] https://t.co/jlFyGtrOKL https://t.co/PK3OV9alrw AB21: 8TF@FP64/32TF@FP32 TSMC 12nm #RISCV @AmpereComputing MUG [----] https://t.co/ixKwdLWaV5 K-AB21 ARM HPC UG"
X Link 2025-12-24T06:05Z [----] followers, [---] engagements
"= DoD RAMP-C Phase [--] Apr [--] [----] Prototype on Intel 18A in [----] Jul [--] [----] RAMP-C customers NVIDIA QCOM Microsoft . Intel NDIA Aug [--] [----] Reuters Dec [--] [----] 🤔🤔🤔 https://www.reuters.com/world/us/how-silicon-valley-dealmaker-charmed-trump-gave-intel-lifeline-2025-12-24/ https://x.com/IanCutress/status/2003969925602271236 https://ndia.dtic.mil/wp-content/uploads/2023/eti/Brinker.pdf https://x.com/ogawa_tter/status/1782618702846058821 Re NVIDIA no longer making test chips on 18A. IT WAS A GOVERNMENT PROGRAM to allow potential partners to build designs using PDKs. It was never a committment"
X Link 2025-12-25T01:44Z [----] followers, [---] engagements
"= Groq and NVIDIA Enter Non-Exclusive Inference Technology Licensing Agreement Dec [--] [----] J. Ross S. Madra & other members of the Groq team will join NVIDIA Groq will continue w/ S. Edwards Deterministic Memory (HBM) D. Abts https://x.com/ogawa_tter/status/1968963891481780677 https://groq.com/newsroom/groq-and-nvidia-enter-non-exclusive-inference-technology-licensing-agreement-to-accelerate-ai-inference-at-global-scale = Groq Raises $750M as Inference Demand Surges Sep [--] [----] https://t.co/VocPTQTweu "Groq next gen chip on Samsung SF4X is going to use 3D stacked SRAM" Oct 2024"
X Link 2025-12-25T02:23Z [----] followers, [----] engagements
"= Optical cross-connect (MEMS OXC) Huawei (CN: May 2024) (CN: Mar 2024) Micromirror Chips (CN: Feb 2024) (CN: Jan 2024) OXC Jul [--] Huawei OXC https://x.com/ogawa_tter/status/1985550170873122966 https://patents.google.com/patent/WO2025152651A1/en https://patents.google.com/patent/WO2025161698A1/en https://patents.google.com/patent/WO2025200814A1/en https://patents.google.com/patent/WO2025246528A1/en = Research Progress of High-Degree OXC Technologies Invited Jul [--] [----] https://t.co/qfLXQNnIK4 Huawei OptiXtrans DC808 Sep [--] [----] https://t.co/ueLHFQvKfa Special Prize at Interop Tokyo [----] Jun"
X Link 2025-12-26T08:31Z [----] followers, [---] engagements
"= Groq & NVIDIA Enter Non-Exclusive Inference Tech Licensing Agreement Dec [--] [----] J. Ross S. Madra & other members of the Groq team will join NVIDIA Dense packaging: HBM Stack Deterministic HBM D. Abts https://research.nvidia.com/person/dennis-abts https://x.com/ogawa_tter/status/1968963891481780677 https://x.com/ogawa_tter/status/2004015056871420165 = Groq Raises $750M as Inference Demand Surges Sep [--] [----] https://t.co/VocPTQTweu "Groq next gen chip on Samsung SF4X is going to use 3D stacked SRAM" Oct [----] https://t.co/zhxAJFh9wX Die-to-Die Dense Packaging (P: Feb 2022)"
X Link 2025-12-26T22:13Z [----] followers, [---] engagements
"= "Semiconductor Packaging Structure and Electronic Device" Huawei Appl (CN: Apr 2024) Stacked DRAM (400) without TSV Appls DRAM (CN: Nov 2021) Stacked DRAM (CN: Dec 2021) Proprietary HBM: HiBL / HiZQ HUAWEI CONNECT [----] Sep [--] https://x.com/ogawa_tter/status/1970494560678879673 https://patents.google.com/patent/WO2025222874A1/en = Huawei Appls DRAM (CN: Nov 2021) https://t.co/GJTecgmYes to reduce an area & power consumption Stacked DRAM (CN: Dec 2021) https://t.co/tz8dSLnjbu [--] Storage Layers Good DRAM Die: 90% Yield: 78.57% (65.61%) HUAWEI CONNECT [----] Sep [--] https://t.co/7yZaSuJsPB"
X Link 2025-12-27T14:28Z [----] followers, [----] engagements
"= "LLM Inference Scheduling: A Survey of Techniques Frameworks and Trade-offs" Huawei Canada/China Simon Fraser U. Tsinghua U. U. of British Columbia TechRxiv Nov [--] [----] Two core architectural layers: Control Plane Data Plane [--] pp [---] ref https://www.techrxiv.org/users/994660/articles/1355915-llm-inference-scheduling-a-survey-of-techniques-frameworks-and-trade-offs https://www.techrxiv.org/users/994660/articles/1355915-llm-inference-scheduling-a-survey-of-techniques-frameworks-and-trade-offs"
X Link 2025-12-29T05:57Z [----] followers, [---] engagements
"= Jul [--] [----] "An Outlook for New Homogeneous Computing System" Huawei Fellow Oct [----] Unified Bus APNet [----] Sep [--] [----] UB-Mesh HC25 SIMT / SIMD https://x.com/ogawa_tter/status/1997191844665630983 https://x.com/ogawa_tter/status/1996559876022247671 https://x.com/ogawa_tter/status/1969593294117355902 https://x.com/ogawa_tter/status/1912038192943170021 https://mp.weixin.qq.com/s/iqRXwdHjyMI7egMrXEIfFw = Graphics Processing Device and Mode Switching Method thereof Huawei Appl (CN: Feb 2023) https://t.co/NWolxy4UcQ Switching between SIMT and SIMD "Matrix Multiplier" US: May [----] (CN: Dec 2017)"
X Link 2025-12-29T06:52Z [----] followers, [---] engagements
"= Huawei Appls 3D SRAM (CN: Sep 2023) 3D 2T1C FeRAM Buffer (CN: Sep 2022) 3D 2T0C/2T2C FeRAM (CN: Aug 2022) FeRAM (CN May/Apr 2022) Stacked DRAM (CN Apr 2024) https://x.com/ogawa_tter/status/2004922428020183375 https://x.com/ogawa_tter/status/1929844531157258464 https://patents.google.com/patent/US20250185254A1/en https://patents.google.com/patent/CN117750776A/en https://patents.google.com/patent/WO2025055404A1/en = "Semiconductor Packaging Structure and Electronic Device" Huawei Appl (CN: Apr 2024) https://t.co/iZuCutuRbj Stacked DRAM (400) without TSV Appls https://t.co/PrbrXhve4d DRAM (CN:"
X Link 2025-12-30T09:20Z [----] followers, [---] engagements
"= "NVIDIA spends $5B on Intel bailout instantly gets $2.5B richer" Dec [--] [----] closed Monday at $36.68 Intel FROM-8-K Dec [--] (26) Sep [--] NVIDIA to invest $5 billion in Intel common stock $23.28/share https://x.com/ogawa_tter/status/1968660914803978356 https://www.intc.com/filings-reports/all-sec-filings/content/0000050863-25-000204/0000050863-25-000204.pdf https://www.theregister.com/2025/12/29/nvidia_intel_5_billion/ = "Intel and NVIDIA to Jointly Develop AI Infrastructure and Personal Computing Products" Sep [--] [----] https://t.co/GZuOacjYEj NVIDIA to invest $5B in Intel common stock will"
X Link 2025-12-30T13:43Z [----] followers, [---] engagements
"= Enabling Dataflow Execution on GPUs w/ Spatial Pipelines NVIDIA ACM TACO Dec [--] [----] K. Sankaralingam Sep 21- Principal Research Scientist NVIDIA Aug 17- Prof UW Madison SimpleMachines (-May 21) https://x.com/ogawa_tter/status/1602458264226045952 https://karu.sites.cs.wisc.edu/wiki/index.php https://dl.acm.org/doi/10.1145/3777466 = SimpleMachines Inc. https://t.co/djds2fJYXe What's . Karthikeyan Sankaralingam https://t.co/xmryTor5vw "founded a chip startup SimpleMachines." K. Sankaralingam et al. ISCA [----] https://t.co/8QWGUpi6x9 Oct [--] [----] https://t.co/Nj60b306AQ https://t.co/s0tYjl99em"
X Link 2025-12-31T12:48Z [----] followers, [---] engagements
""Hardware Acceleration for Neural Networks: A Comprehensive Survey" ASU arXiv Dec [--] [----] [--] pp [--] ref S. Gupta = Zhonghao Xinying () 2nd-gen (7nm) in [----] Dec [--] : VLIW AI Chip (TPU) https://x.com/ogawa_tter/status/2005984006467379568 https://scholar.google.com/citationsuser=U9bcQkMAAAAJ&view_op=list_works&sortby=pubdate https://arxiv.org/abs/2512.23914 = Dialogue with Tech Innovators Dec [--] [----] https://t.co/I4PjNVea64 Yanggong Yifan Founder & CEO of 1st-gen [----] 2nd-gen (7nm) in [----] Local alternative to Google's TPU SCMP Nov [--] https://t.co/pthfwmLzVY https://t.co/4fVbKaZ2jQ VLIW AI Chip"
X Link 2026-01-02T07:57Z [----] followers, [----] engagements
"Tech Trends in Computing HW & their Impacts on High-Performance Sci Computing Part I: General-Purpose Processors & HW Accelerators UT Austin & ExxonMobil IJHPCA Dec [--] [----] (OnlineFirst) [--] pp [---] ref = Huawei Oct [----] https://x.com/ogawa_tter/status/2005532420033155255 https://journals.sagepub.com/doi/full/10.1177/10943420251348799 = Jul [--] [----] https://t.co/ZDi6UxGD7m "An Outlook for New Homogeneous Computing System" Huawei Fellow Oct [----] Unified Bus APNet [----] https://t.co/eYXz5ulaJU Sep [--] [----] https://t.co/LszYpCM0Xe UB-Mesh HC25 https://t.co/NXZRshSxio https://t.co/aZhtX280NL SIMT /"
X Link 2026-01-02T13:40Z [----] followers, [---] engagements
"= Baidu Announces Proposed Spin-off and Separate Listing of Kunlunxin () Jan [--] [----] Reuters valued it at $3B Dec [--] [----] Baidu World [----] Nov [--] https://x.com/ogawa_tter/status/1989203706043330700 https://www.reuters.com/world/asia-pacific/baidus-ai-chip-arm-kunlunxin-files-confidentially-hong-kong-listing-2026-01-01/ https://www.kunlunxin.com/ https://asia.cdn.euroland.com/press-releases-attachments./3955357/HKEX-EPS_20260102_11974277_0.PDF https://baidu.gcs-web.com/node/14381/pdf = (Baidu World) [----] Nov [--] https://t.co/0a4vhPeoMA Live https://t.co/L51mmUhB1V Kunlun () M100 (Early 2026)/M300"
X Link 2026-01-04T03:33Z [----] followers, [---] engagements
"= "KernelEvolve: Scaling Agentic Kernel Coding for Heterogeneous AI Accelerators at Meta" Dec [--] [----] arXiv (Jan [--] 2026) NVIDIA A100 / H100 AMD MI300 / MI350 MTIA v3 (Metas latest-gen AI accelerators) / v2i MTIA v2i ISCA [----] https://x.com/ogawa_tter/status/1936349816545763486 https://arxiv.org/abs/2512.23236 = "Meta's Second Generation AI Chip: Model-Chip Co-Design and Productionization Experiences" Industry Track ISCA [----] https://t.co/N34ujcpDh9 MTIA 2i Hot Chips [----] https://t.co/6xvaGCWAd9 https://t.co/m5Srr8wVXm Andes #RISCV Day Tokyo Feb [--] https://t.co/JWLJG79zKS Arteris"
X Link 2026-01-04T09:09Z [----] followers, [---] engagements
"= Biren Technology () One of China's "Four Little Dragons" of GPUs has entered the capital market Jan [--] [----] [----] HK GLOBAL OFFERING Dec [--] [----] (663 pp) Jul [--] [----] https://x.com/ogawa_tter/status/1950816787647570069 https://www1.hkexnews.hk/listedco/listconews/sehk/2025/1222/2025122200019.pdf https://www.hkex.com.hk/Market-Data/Securities-Prices/Equities/Equities-Quotesym=6082&sc_lang=zh-HK https://www.c114.com.cn/news/16/a1303403.html = "Communication Board Topology and Communication System" & (Xizhi Tech) Appl (CN: Feb 2025) https://t.co/t2f2upJhje GPUs are connected by Switch Chips/dOCSs"
X Link 2026-01-05T09:28Z [----] followers, [---] engagements
"= CXMT ()'s IPO application on the has been accepted Dec [--] [----] (Previewed) Prospectus Dec [--] (365 pp) no "HBM" word (Q&A) HBM 🤔🤔 https://static.sse.com.cn/stock/disclosure/announcement/c/202512/002170_20251230_XS2U.pdf https://static.sse.com.cn/stock/disclosure/announcement/c/202512/002170_20251230_B8QS.pdf https://www.sse.com.cn/listing/renewal/ipo/index_listing_detail.shtmlauditId=2170 https://eet-china.com/news/202512314093.html https://static.sse.com.cn/stock/disclosure/announcement/c/202512/002170_20251230_XS2U.pdf"
X Link 2026-01-05T14:59Z [----] followers, [---] engagements
"= "Hybrid Computing Network Topologies for Workload Execution" Brian P. Towles Yixin Luo Google (Mar 2024) can flexibly allocate bandwidth etc. "Enhanced Reconfigurable Interconnect Network" Granted: Jun [----] (P: Oct 2020) https://x.com/ogawa_tter/status/1619081874596585472 https://patents.google.com/patent/US20250284529A1/en = "Enhanced Reconfigurable Interconnect Network" Google Patent Appl Apr [----] (Dec 2020) https://t.co/fj1IEKREU1 Reconfigurable Interconnect Fabric for Scalability Optimized Network Bandwidth Optimized Network Latency Optimized Network TPUv4 Oct 2022"
X Link 2026-01-06T08:31Z [----] followers, [---] engagements
"@miyox @yyasuda @ebiken @O_FiberLady 9NVIDIA CEO/ Enfabrica Legal Events ENTITY STATUS OF PATENT OWNER : SMALL ENTITY 2025-10-30: LARGE ENTITY https://x.com/ogawa_tter/status/1620466846888775686 https://x.com/ogawa_tter/status/1968860255934173256 = Enfabrica (Jun 2022) Transparent Remote Memory Access ovr NW Protocol https://t.co/wRY8pBxT9O Multi-Plane /-Protocol Memory Switch Fabric w/ Configurable Transport https://t.co/aBB1cZ6WGa Server Fabric Adapter (Jan 2022) https://t.co/KU3T1lTSyv Jun [----] https://t.co/8rzxBuZnFU https://t.co/jfveOb2KGO"
X Link 2026-01-06T10:58Z [----] followers, [---] engagements
"= CES [----] Live with J. Huang Feb [--] Financial Analyst Q&A News Rubin - Six New Chips BF-4 Inference Context Memory Storage Platform https://nvidianews.nvidia.com/news/nvidia-bluefield-4-powers-new-class-of-ai-native-storage-infrastructure-for-the-next-frontier-of-ai https://nvidianews.nvidia.com/news/rubin-platform-ai-supercomputer https://nvidianews.nvidia.com/online-press-kit/ces-2026-news https://video.ibm.com/recorded/134654723 https://s201.q4cdn.com/141608511/files/doc_downloads/2026/01/05/JHH-CES-2026-FINAL.pdf https://www.youtube.com/watchv=0NBILspM4c4&t=6206s"
X Link 2026-01-07T07:38Z [----] followers, [--] engagements
"= Tenstorrent: First Gen Compact AI Accelerator Device Jan [--] [----] Partnership with Razer Wormhole tech Compact & Modular AI accelerator brings powerful GenAI capabilities to any Thunderbolt [--] or [--] equipped system up to [--] devices https://www.razer.com/newsroom/company-news/razer-at-ces-2026/ https://tenstorrent.com/vision/tenstorrent-unveils-first-gen-compact-ai-accelerator-device https://www.razer.com/newsroom/company-news/razer-at-ces-2026/ https://tenstorrent.com/vision/tenstorrent-unveils-first-gen-compact-ai-accelerator-device"
X Link 2026-01-07T07:54Z [----] followers, [---] engagements
"= AMD CES [----] Jan [--] Yottascale Era Press Kit Financial AD Nov [--] UALoE HPE Dec [--] Radiation Effects Oct [--] https://x.com/ogawa_tter/status/2001567210595324132 https://x.com/ogawa_tter/status/1996037944759746638 https://x.com/ogawa_tter/status/1988608702807691427 https://www.amd.com/en/newsroom/press-kits/ces-2026.html https://www.amd.com/en/blogs/2026/from-ces-2026-to-yottaflops-why-the-amd-keynote.html https://www.amd.com/content/dam/amd/en/documents/corporate/events/amd-ces-2026-distribution-deck.pdf https://www.youtube.com/watchv=UbfAhFxDomE "Radiation Effects in FPGAs and SoCs: Adapting"
X Link 2026-01-07T08:53Z [----] followers, [---] engagements
"= NVIDIA CES [----] J. Huang Feb [--] Financial Analyst Q&A News Rubin - Six New Chips BF-4 Inference Context Memory Storage Platform https://nvidianews.nvidia.com/news/nvidia-bluefield-4-powers-new-class-of-ai-native-storage-infrastructure-for-the-next-frontier-of-ai https://nvidianews.nvidia.com/news/rubin-platform-ai-supercomputer https://nvidianews.nvidia.com/online-press-kit/ces-2026-news https://video.ibm.com/recorded/134654723 https://s201.q4cdn.com/141608511/files/doc_downloads/2026/01/05/JHH-CES-2026-FINAL.pdf https://youtube.com/watchv=0NBILspM4c4&t=6206s"
X Link 2026-01-07T09:04Z [----] followers, [--] engagements
"= NVIDIA CES [----] J. Huang Feb [--] Financial Analyst Q&A News Rubin - Six New Chips BF-4 Inference Context Memory Storage Platform https://nvidianews.nvidia.com/news/nvidia-bluefield-4-powers-new-class-of-ai-native-storage-infrastructure-for-the-next-frontier-of-ai https://nvidianews.nvidia.com/news/rubin-platform-ai-supercomputer https://nvidianews.nvidia.com/online-press-kit/ces-2026-news https://video.ibm.com/recorded/134654723 https://s201.q4cdn.com/141608511/files/doc_downloads/2026/01/05/JHH-CES-2026-FINAL.pdf https://youtube.com/watchv=0NBILspM4c4&t=6206s"
X Link 2026-01-07T09:09Z [----] followers, [---] engagements
"= Intel CES [----] Core Ultra Series [--] Launch Event Jan [--] [----] Lip-Bu Tan Series [--] Deep Dive (88 pp) Press Kit Series [--] Debut as First Built on Intel 18A https://newsroom.intel.com/client-computing/ces-2026-intel-core-ultra-series-3-debut-first-built-on-intel-18a https://newsroom.intel.com/press-kit/press-kit-intel-at-ces-2026 https://download.intel.com/newsroom/2026/CES2026/Intel-CES2026-Core-Ultra-Series-3-Press-Deck.pdf https://download.intel.com/newsroom/2026/CES2026/Intel-CES%202026-Core-Ultra-Series-3-Keynote-010526.pdf https://www.youtube.com/watchv=KlIlFt2Fj1c"
X Link 2026-01-07T11:21Z [----] followers, [---] engagements
"= NVIDIA Inference Context Memory Storage Platform CES [----] Jan [--] BlueField-4 (Eth RDMA)-powered Flash-based G3.5 context memory tier Dynamo KVBM Oct [----] Enfabrica G2.5 Memory https://x.com/ogawa_tter/status/1968860255934173256 https://www.youtube.com/watchv=f5O3BrGwZVE https://developer.nvidia.com/blog/introducing-nvidia-bluefield-4-powered-inference-context-memory-storage-platform-for-the-next-frontier-of-ai/ https://x.com/ogawa_tter/status/2008828254308970667 = NVIDIA just spent over $900 million to hire @enfabrica_ CEO Rochan Sankar & other employees license AI startups technology Sep 18"
X Link 2026-01-08T06:54Z [----] followers, [----] engagements
"NVIDIA CES [----] Inference Context Memory Storage Platform BlueField-4 Eth RDMA Flash-based G3.5 context memory tier G1: GPU HBM G2: System RAM G3: Local SSDs G4: Shared storage Enfabrica G2.5 Mem Pool https://x.com/ogawa_tter/status/2009156839439777817 = NVIDIA Inference Context Memory Storage Platform CES [----] https://t.co/wMoR85lHT7 Jan [--] https://t.co/9GQsVVcXHp BlueField-4 (Eth RDMA)-powered Flash-based G3.5 context memory tier Dynamo KVBM Oct [----] https://t.co/UKOs1h0x64 Enfabrica https://t.co/jIGyT2BHx2 G2.5 Memory https://t.co/KwWwuvawvl"
X Link 2026-01-08T07:35Z [----] followers, [---] engagements
"= "Challenges and Research Directions for Large Language Model Inference Hardware" X. Ma David Patterson Google arXiv Jan [--] [----] [--] Research Opportunities to Re-Think HBF NMC 3D Stacking LL Interconnect https://x.com/ogawa_tter/status/1991789595214893165 https://x.com/ogawa_tter/status/1949068347565547896 https://arxiv.org/abs/2601.05047 = "Computer Architecture [---] & Its Future" David Patterson WS SimonsInst Sep [--] [----] https://t.co/pe4kVzqmHa TPU [--] yrs Aug [--] [----] https://t.co/fZJjK48sry [----] - [--] [----] https://t.co/MVZntG5mpc Emissions https://t.co/1y6H9cbs8Z How to Have a Bad"
X Link 2026-01-09T03:53Z [----] followers, [---] engagements
"= State-of-the-Art Analysis and Perspectives of China HPC Development in . [----] [----] [----] [----] [----] [----] http://www.jfdc.cnic.cn/CN/10.11871/jfdc.issn.2096-742X.2020.06.001 http://www.jfdc.cnic.cn/CN/10.11871/jfdc.10-1649.2021.06.007 http://www.jfdc.cnic.cn/CN/10.11871/jfdc.issn.2096-742X.2022.06.001 http://www.jfdc.cnic.cn/CN/10.11871/jfdc.issn.2096-742X.2023.06.001 http://www.jfdc.cnic.cn/CN/10.11871/jfdc.issn.2096-742X.2024.06.001 http://www.jfdc.cnic.cn/CN/10.11871/jfdc.issn.2096-742X.2025.06.009 http://www.jfdc.cnic.cn/CN/10.11871/jfdc.issn.2096-742X.2020.06.001"
X Link 2026-01-09T13:49Z [----] followers, [---] engagements
"= Tech Trends in Computing HW & their Impacts on High-Performance Sci Computing Part II: Memory systems Interconnects and System integration UT Austin & ExxonMobil IJHPCA Jan [--] [----] (OnlineFirst) [--] pp [---] ref Part I Dec [--] https://x.com/ogawa_tter/status/2007084619737563357 https://journals.sagepub.com/doi/full/10.1177/10943420251347461 Tech Trends in Computing HW & their Impacts on High-Performance Sci Computing Part I: General-Purpose Processors & HW Accelerators UT Austin & ExxonMobil IJHPCA Dec [--] [----] (OnlineFirst) https://t.co/aN24vurNQE [--] pp [---] ref = Huawei Oct 2021"
X Link 2026-01-09T14:46Z [----] followers, [---] engagements
"= Semiconductors: Information on Projects Funded to Strengthen U.S. Supply Chain Dec [--] [----] GAO cancels $45M Oregon State grant for Microfluidics May [--] terminates $285M contract for SMART USA Dec [--] https://x.com/ogawa_tter/status/2001156497288613985 https://x.com/ogawa_tter/status/1971941691994300867 https://www.gao.gov/products/gao-26-107882 = Federal gov terminates $285M contract for SMART USA CHIPS manufacturing institute Dec [--] [----] https://t.co/LeEYy4RuNm 'termination for convenience' Executive Director https://t.co/3ztuy4COqT ORAU Webinar Oct [--] https://t.co/hL1DxWApT3 Feb 19"
X Link 2026-01-10T03:05Z [----] followers, [---] engagements
"= "What is Groq-NVIDIA Deal Really About" Jan [--] [----] @sallywf "may include a desire to stop one of these hyperscale customers from buying Groq " Silicon Tiler Ecosystem Dense packag Deterministic HBM https://x.com/ogawa_tter/status/2004677049685008484 https://x.com/ogawa_tter/status/1770734477066608744 https://www.eetimes.com/what-is-groq-nvidia-deal-really-about/ = Groq & NVIDIA Enter Non-Exclusive Inference Tech Licensing Agreement Dec [--] [----] https://t.co/mGkLL3xafG J. Ross S. Madra & other members of the Groq team will join NVIDIA https://t.co/QqlSILw38M Dense packaging: HBM Stack"
X Link 2026-01-10T08:23Z [----] followers, [---] engagements
"= Xiaoyu Ma David Patterson. Google arXiv Jan [--] [----] "Tree-based Network Architecture for Accelerating Machine Learning Collective Operations" Nov [----] (Jan 2024) LLM with 3D-DRAM Chiplets (P: Jun 2023) https://x.com/ogawa_tter/status/1897205643519684768 https://patents.google.com/patent/US12481608B2/en https://x.com/ogawa_tter/status/2009473548679815209 = "Serving Large Language Models with 3D-DRAM Chiplets" Xiaoyu Ma David Patterson Google (P: Jun 2023) https://t.co/OjCeLg4HBd Compute-memory chiplets using small low-power comp dies with low arithmetic intensity (32 TF 40W TDP) NVIDIA"
X Link 2026-01-11T08:16Z [----] followers, [---] engagements
"= Interview w/ Mitesh Agarwal CEO of Positron Dec [--] [----] Journey to Redefine AI HW Two core innovations 90+% memory utilization Multi-model concurrency Taping out an ASIC Agilex-7M FPGA $51.6M Jul [--] https://x.com/ogawa_tter/status/1950213984264868004 https://x.com/ogawa_tter/status/1978738177457438975 https://eqvista.com/interview-with-mitesh-agarwal-positron/ = Positron AI Secures $51.6M in Oversubscribed Series A to Accelerate Inference-Optimized Hardware Jul [--] [----] https://t.co/GEJjxbaMt6 Atlas: First-gen FPGA-based $23.5M seed round Feb [--] https://t.co/cGsG7ssFiY Agilex-7M"
X Link 2026-01-11T15:01Z [----] followers, [---] engagements
"= "Performance efficiency and cost analysis of wafer-scale AI accelerators vs. single-chip GPUs" UC Riverside Review Device Oct [--] [----] Cerebras WSE-3 Tesla Dojo (SoW) NVIDIA H100 Google TPU v6e [--] pp [---] ref M. Ozkan https://scholar.google.com/citationsuser=zE75wgsAAAAJ&view_op=list_works&sortby=pubdate https://www.sciencedirect.com/science/article/pii/S2666998625001474 https://scholar.google.com/citationsuser=zE75wgsAAAAJ&view_op=list_works&sortby=pubdate https://www.sciencedirect.com/science/article/pii/S2666998625001474"
X Link 2026-01-12T05:08Z [----] followers, [---] engagements
"= IBM Z Tech Bytes Session Proceedings for May [----] May 15-15 IBM Z Performance Hot Topics (31 pp) Tech Overview (64 pp) Hot Chips [----] z17 Processor Module https://x.com/ogawa_tter/status/1918253608635888043 https://x.com/ogawa_tter/status/1911330163939291149 https://www.ibm.com/support/pages/system/files/inline-files/IBM%20z17%20and%20LinuxONE%205%20Technical%20Overview.pdf https://www.ibm.com/support/pages/system/files/inline-files/Tech%20Bytes%20May%202025%20IBM%20Z%20Hot%20Topics.pdf https://www.ibm.com/support/pages/node/7232338 = "The New IBM z17 Telum II Processor Module Cut Open Down"
X Link 2026-01-13T08:37Z [----] followers, [---] engagements
"= NN SoC in AR/VR glasses Meta Appls (Aug 2023) 3D Stacked Hybrid Memory MEMSYS [----] Stacked SRAM Face-to-Face Hybrid Bonding SRAM for AR Aug [----] (Sep 2021) https://patents.google.com/patent/US12068054B2/en https://patents.google.com/patent/US20250054910A1/en https://patents.google.com/patent/US20250054911A1/en https://x.com/ogawa_tter/status/1887752621361926464 https://patents.google.com/patent/US20250056815A1/en = "Characterization and Design of 3D-Stacked Memory for Image Signal Processing on AR/VR Devices" Meta MEMSYS [----] https://t.co/ZdzZAooBM5 3D-Stacked Hybrid Memory ISP solution"
X Link 2026-01-13T12:48Z [----] followers, [---] engagements
"= IBM Z Tech Bytes Session Proceedings for May [----] May 5-15 IBM Z Performance Hot Topics (31 pp) Tech Overview (64 pp) Hot Chips [----] z17 Processor Module https://x.com/ogawa_tter/status/1918253608635888043 https://x.com/ogawa_tter/status/1911330163939291149 https://www.ibm.com/support/pages/system/files/inline-files/IBM%20z17%20and%20LinuxONE%205%20Technical%20Overview.pdf https://www.ibm.com/support/pages/system/files/inline-files/Tech%20Bytes%20May%202025%20IBM%20Z%20Hot%20Topics.pdf https://www.ibm.com/support/pages/node/7232338 = "The New IBM z17 Telum II Processor Module Cut Open Down"
X Link 2026-01-14T12:35Z [----] followers, [---] engagements
"= "SmartBlock: Adaptive Block Floating Point Quantization for Efficient DNN Acceleration" NUDT ICPP [----] Sep [--] Systolic Array: Unified PE Arch Block Size: Dynamic Google (P: Apr 2024) Systolic Array Block Data Structured Sparse https://x.com/ogawa_tter/status/1997305240270618950 https://dl.acm.org/doi/10.1145/3754598.3754660 = "Structured Sparse Matrix Acceleration In Systolic Arrays " Google Appls (P: Apr 2024) https://t.co/4lEIYrhPPf Dense and Structured Fine-Grained Sparse Block Data Type TPUv7: Ironwood Hot Chips [----] https://t.co/gqG74DgNLB Self-Repairable Chip (Apr 2024) & BIST (Feb"
X Link 2026-01-14T14:39Z [----] followers, [---] engagements
"= "High-Performance Computing and Responsibly Reckless Algorithms" Jack Dongarra Keynote ICPP [----] Sep [--] "Published" China HPC Top [---] 2025-20 Ozaki Scheme NVIDIA SCA/HPCAsia [----] https://x.com/ogawa_tter/status/2000530683064938507 https://x.com/ogawa_tter/status/2009623488596988004 https://ssl.linklings.net/conferences/icpp/icpp2025_program/views/includes/files/evt_copy101s1.pdf = "Guaranteed DGEMM Accuracy While Using Reduced Precision Tensor Cores Through Extensions of the Ozaki Scheme" NVIDIA SCA/HPCAsia [----] Jan [--] https://t.co/KEN71PYZZP Ozaki-II scheme Dec [--] https://t.co/XqAeyqdVrr Oct"
X Link 2026-01-14T15:20Z [----] followers, [---] engagements
"= "OpenAI partners with Cerebras" Jan [--] [----] to add 750MW of ultra low-latency AI compute to our platform online in multiple tranches through [----] Wafer-Scale Systems Carbon Perspective Aug [----] Review Oct [----] [--] pp [---] ref https://x.com/ogawa_tter/status/2011071479715414082 https://openai.com/index/cerebras-partnership/ = "Wafer-Scale Systems: A Carbon Perspective" ACM SIGEnergy Energy Informatics Review Aug [--] [----] https://t.co/uIYXOkOtir Cerebras CS-3 vs H100 systems Performance Efficiency & Cost analysis of Wafer-Scale AI Accelerators vs. Single-Chip GPUs Oct [--] https://t.co/yRytINwLQH"
X Link 2026-01-15T02:17Z [----] followers, [---] engagements
"= AMD Appls (Jun 2024) "Balanced Latency Stacked Cache" [--] MB L2M cache: [--] cycle latency (= 14) "Apparatus and System for Multi-Purpose Macros for Vertical Die Interconnects" disclosed macro cells can be used in stacking caches https://patents.google.com/patent/US20250391815A1/en https://patents.google.com/patent/US20260003794A1/en https://patents.google.com/patent/US20250391815A1/en https://patents.google.com/patent/US20260003794A1/en"
X Link 2026-01-15T07:03Z [----] followers, [---] engagements
"= "A Path to 448Gbps" Tad Hofmeister Google Keynote OIF 448Gbps Signaling for AI WS Apr [--] [----] Report PAM6/8 are more realistic paths [---] G/Lane May [--] / Apr [--] CPC Sep [--] https://x.com/ogawa_tter/status/1976861738051289284 https://x.com/ogawa_tter/status/1930633833541742644 https://convergedigest.com/oif-keynote-google-urges-swift-move-to-448g-for-ai-clusters/ https://www.oiforum.com/wp-content/uploads/Tad-A-Path-to-448Gbps.pdf = "Revolutionizing Rackscale Connectivity using Co-Packaged Copper and Optics" Samtec & Marvell [----] ECOC Product Focus Sep [--] https://t.co/SAZM4HeUBE Marvell"
X Link 2026-01-15T14:39Z [----] followers, [----] engagements
"= "AI Chip Startup Etched Raises $500 Million to Take on NVIDIA" Jan [--] [----] Bloomberg according to people familiar with the matter Jun [--] [----] Page does not exist🤔 G. Uberti Co-Founder & CEO Feb [--] https://x.com/ogawa_tter/status/1805775421235151153 https://www.etched.com/ https://finance.yahoo.com/news/ai-chip-startup-etched-raises-205348019.html = "@Etched is Making the Biggest Bet in AI" Jun [--] [----] https://t.co/EOfmQAMeq1 Sohu the worlds first transformer ASIC can only run one algorithm One 8x Sohu server replaces 160x H100 GPUs. TSMC 4nm Gavin Uberti Co-Founder & CEO Etched Feb 28"
X Link 2026-01-16T02:50Z [----] followers, [---] engagements
"= "Trump administration sets GPU export rules that put Chinese buyers at the back of the queue" Jan [--] [----] Revision to License Review Policy for Advanced Computing Commodities Jan [--] covered by this rule are the NVIDIA H200 . https://www.federalregister.gov/documents/2026/01/15/2026-00789/revision-to-license-review-policy-for-advanced-computing-commodities https://www.theregister.com/2026/01/14/us_gpu_export_rules/ https://www.federalregister.gov/documents/2026/01/15/2026-00789/revision-to-license-review-policy-for-advanced-computing-commodities"
X Link 2026-01-16T07:09Z [----] followers, [---] engagements
"= Fact Sheet: Restoring American Semiconductor Manufacturing Leadership Through an Agreement on Trade & Investment with Taiwan Jan [--] [----] U.S. Dep of Commerce Direct Investments: $250+ billion Additional Investments: $250+ billion Industrial Parks https://www.commerce.gov/news/fact-sheets/2026/01/fact-sheet-restoring-american-semiconductor-manufacturing-leadership https://www.commerce.gov/news/fact-sheets/2026/01/fact-sheet-restoring-american-semiconductor-manufacturing-leadership"
X Link 2026-01-16T10:41Z [----] followers, [---] engagements
"= MatX (P: May 2024) Re-Rounding in Integrated Circuit for Variance Reduction in AI Ops R. Pope Acyclic Arch R. Pope & M. Gunter Mar [--] [----] $100M Series A https://x.com/reinerpope/status/1899486276233146530 https://x.com/ogawa_tter/status/1774422751253823814 https://patents.google.com/patent/US20250355715A1/en https://patents.google.com/patent/US20250355711A1/en https://matx.com/ MatX is designing chips and systems to 10x the computing power for the worlds largest AI workloads. Today we are pleased to announce the closing of a $100M Series A funding round led by @sparkcapital with"
X Link 2026-01-16T13:18Z [----] followers, [---] engagements
"= Breaking the Memory Wall: Scaling AI Inference with Innovative Memory Fanout Arch On-Demand (Nov [--] 2025) Credo Nov [--] OmniConnect 112G VSR Weaver Memory Fanout Gearbox M. Agrawal CEO Positron AI Dec [--] https://x.com/ogawa_tter/status/2010366527288262767 https://investors.credosemi.com/news-events/news/news-details/2025/Credo-Unveils-Industrys-First-Memory-Fanout-Gearbox-for-Scalable-High-Bandwidth-AI-Inference/default.aspx https://event.on24.com/wcc/r/5135492/10AE35BE7833F4F6B7C17C80FFFCA319 = Interview w/ Mitesh Agarwal CEO of Positron Dec [--] [----] https://t.co/32BRMNr4zG Journey to"
X Link 2026-01-17T02:05Z [----] followers, [---] engagements
"= "HyGain: High-performance Energy-efficient Hybrid Gain Cell-based Cache Hierarchy" ACM TACO Mar [--] [----] 2T Gain Cell Cache Memory OS GC Memory HSP Wong Stanford with TSMC NVIDIA IEDM [----] May [----] https://x.com/ogawa_tter/status/1994399791426703746 https://x.com/ogawa_tter/status/1996939609570529777 https://dl.acm.org/doi/10.1145/3572839 = (Oxide Semiconductor) Gain Cell Memory: Technology Integration & Memory Access H.-S. Philip Wong Stanford Stanford DAM Kickoff WS May [--] [----] https://t.co/kYNvY4ybUC Trans on Electron Devices Apr [--] https://t.co/BWIUxIO1l8 RAAAM Memory"
X Link 2026-01-18T14:03Z [----] followers, [---] engagements
"= "SK Hynix's $13B packaging facility promises more HBM for the AI bubble" Jan [--] [----] [----] Market Outlook "Focus on the HBM-led memory supercycle" Jan [--] SK HBM Market Share SK HBM4 Tech Leadership https://news.skhynix.co.kr/2026-market-outlook/ https://news.skhynix.co.kr/fact-02/ https://www.theregister.com/2026/01/13/sk_hbm_packaging/ https://news.skhynix.co.kr/2026-market-outlook/ https://news.skhynix.co.kr/fact-02/ https://www.theregister.com/2026/01/13/sk_hbm_packaging/"
X Link 2026-01-19T08:34Z [----] followers, [---] engagements
"= "Optimization and Benchmarking of Monolithically Stackable Gain Cell Memory for Last-Level Cache" GaTech & UV arXiv Mar [--] [----] 7/3 nm HD SRAM 64/128 MB AOS-based 2T-GC [---] MB Review Jan [----] Stanford TSMC NVIDIA: IEDM25 https://x.com/ogawa_tter/status/2010608213436289284 https://arxiv.org/abs/2503.06304 = "Oxide Semiconductor for Advanced Memory Architectures: Atomic Layer Deposition Key Requirement and Challenges" Hanyang Univ Review Nano-Micro Letters Jan [--] [----] https://t.co/qTyZdUtkyY [--] pp [---] ref Huawei Appls https://t.co/sgEIsl93RA Stanford https://t.co/DOugYZznRW"
X Link 2026-01-19T14:49Z [----] followers, [----] engagements
"= GPU: HBM + (LP) DDR DRAM NVIDIA Patent Oct [----] (Apr 2021) ASSYST [----] with GaTech H100 HBM + Tensor Offloading-Memory AMD "Integrated Memory Controller Circuitry" (P: Sep 2023) MI455X HBM4 + LPDDR5 https://x.com/SemiAnalysis_/status/2013310524109328580 https://patents.google.com/patent/US20250098181A1/en https://x.com/ogawa_tter/status/1722957240334946367 In AMD's CES [----] keynote we spotted [--] LPDDR5 DRAM packages beside the MI455XAMD has not said anything about this to date but SemiAnalysis has detailed this hybrid HBM4+LPDDR5 memory configuration in our article from last April. This is"
X Link 2026-01-20T05:41Z [----] followers, [---] engagements
"= "Micron breaks ground on humungous NY DRAM fab after beating bats and tree huggers" Jan [--] [----] had been bedeviled by environmental delays celebrates official groundbreaking at NY megafab site Dec [----] https://x.com/ogawa_tter/status/1866651828701761955 https://www.micron.com/us-expansion/new-york https://www.theregister.com/2026/01/16/micron_fab_breaks_ground/ = Dep of Commerce Awards CHIPS Incentives to Micron for Idaho & New York Projects and PMT for Virginia DRAM Project to Secure Domestic Supply of Legacy Memory Dec [--] [----] https://t.co/Y0EDFzrfcm $4.6B & $1.5B in Direct + $65M VA PMT:"
X Link 2026-01-20T06:55Z [----] followers, [---] engagements
"@1p_semicon (4/8 Gain Cell Memory Stanford H.-S. Philip Wong TSMC 2018-20 VP Corporate Res Chief Scientist IEDM [----] NVIDIA GC-Memory RT https://x.com/ogawa_tter/status/1996939609570529777 = Gain Cell Memory IEDM [----] Dec [--] H.-S. Philip Wong Stanford DAM https://t.co/jNi5fUFjbv https://t.co/Zyrfs3WAY9 & TSMC NVIDIA 28-8 "Amp Cell: 3D-Scalable Gain Cell Memory with Decoupled Read Write and Store" 29-5 "Gain Cell Memory Scalability to 5-nm and Beyond" +Purdue https://t.co/Y5SZaAKP15 https://x.com/ogawa_tter/status/1996939609570529777 = Gain Cell Memory IEDM [----] Dec [--] H.-S. Philip Wong"
X Link 2026-01-21T05:28Z [----] followers, [----] engagements
"@1p_semicon (5/8 Preprint (arXiv Mar [--] 2025) Amorphous Oxide Semiconductor-based 2-Transistor Gain Cell Memory (AOS-based 2T-GC Memory) High-Density SRAMLast-Level Cache (64/128 MB) 7nm / 3nm https://x.com/ogawa_tter/status/2013262449739244018 = "Optimization and Benchmarking of Monolithically Stackable Gain Cell Memory for Last-Level Cache" GaTech & UV arXiv Mar [--] [----] https://t.co/kxph41Y1RE 7/3 nm HD SRAM 64/128 MB AOS-based 2T-GC [---] MB Review Jan [----] https://t.co/0szYhdisDf Stanford TSMC NVIDIA: IEDM25 https://t.co/rvxaShp0yf https://x.com/ogawa_tter/status/2013262449739244018 ="
X Link 2026-01-21T05:29Z [----] followers, [----] engagements
"@1p_semicon (6/8 Amorphous Oxide Semicon-based Tr2 H.-S. Philip Wong Oxide Semicon OS for Advanced Memory Arch Review (Nano-Micro Letters Jan [--] 2026) IGZO Memory https://x.com/ogawa_tter/status/2010608213436289284 = "Oxide Semiconductor for Advanced Memory Architectures: Atomic Layer Deposition Key Requirement and Challenges" Hanyang Univ Review Nano-Micro Letters Jan [--] [----] https://t.co/qTyZdUtkyY [--] pp [---] ref Huawei Appls https://t.co/sgEIsl93RA Stanford https://t.co/DOugYZznRW https://t.co/wJDqGcUgxR https://x.com/ogawa_tter/status/2010608213436289284 = "Oxide Semiconductor for Advanced"
X Link 2026-01-21T05:30Z [----] followers, [----] engagements
"= Upscale AI Closes Oversubscribed $200M Series A Jan [--] [----] Tota: $300+M Oct [--] [----] $100M Seed Sep [--] SkyHammer Arch Oct [--] Aug [----] (Apr 2025) to separate Header and Data https://patents.google.com/patent/US12388770B1/en https://upscaleai.com/upscale-ai-unveils-skyhammer-architecture/ https://x.com/ogawa_tter/status/1981691384538316808 https://upscaleai.com/from-100m-seed-to-unicorn-in-months-upscale-ai-closes-oversubscribed-200m-series-a-to-build-the-first-pure-play-ai-networking-company/ = Perform Evaluation of Interconnect Tech for AI Scale-Up Comp: UAL vs UALoE/SUE vs RoCE Upscale AI"
X Link 2026-01-22T03:09Z [----] followers, [---] engagements
"= "VCSEL-based CPO for Scale-Up in A.I. Datacenter. Status and Perspectives" arXiv Jan [--] [----] Mature VCSEL offers the ideal combination M Kohli COHR J Teissier COHR VCSEL COHR https://www.coherent.com/networking/optoelectronic-devices/datacom-vcsel-photodiodes/vcsel-arrays https://scholar.google.com/citationsuser=BeW1B9wAAAAJ&view_op=list_works&sortby=pubdate https://scholar.google.com/citationsuser=kMPss_sAAAAJ&view_op=list_works&sortby=pubdate https://arxiv.org/abs/2601.14342 https://www.coherent.com/networking/optoelectronic-devices/datacom-vcsel-photodiodes/vcsel-arrays"
X Link 2026-01-22T05:44Z [----] followers, [---] engagements
"@kazunori_279 @kumagi @miyox @yyasuda @ebiken @O_FiberLady ( HBF (High Bandwidth Flash) Jul [--] [----] Patterson Sandisk HBF Technical Advisory Board arXiv LLM 3D Stack Chips ML Collective Operations () ( https://x.com/ogawa_tter/status/2010264445755548026 https://x.com/ogawa_tter/status/1949068347565547896 = Xiaoyu Ma David Patterson. Google arXiv Jan [--] [----] https://t.co/lvUhAzkOqo "Tree-based Network Architecture for Accelerating Machine Learning Collective Operations" Nov [----] (Jan 2024) https://t.co/R5vQkMJY1j LLM with 3D-DRAM Chiplets (P: Jun 2023) https://t.co/GURhSDp8uu"
X Link 2026-01-23T05:59Z [----] followers, [---] engagements
"= "How to Write High-Performance Matrix Multiply in NVIDIA CUDA Tile" Jan [--] [----] How to implement matrix multiplication using NVIDIA cuTile Block-level parallel programming mindset Best practices GPU Mode Dec [--] [----] (1:43:20) https://x.com/ogawa_tter/status/2002971559053639796 https://developer.nvidia.com/blog/how-to-write-high-performance-matrix-multiply-in-nvidia-cuda-tile/ = NVIDIA CUDA Tile using cuTile and CUDA Tile IR to write portable high-performance GPU Kernels @JokerEph & @roeschinc GPU Mode Dec [--] [----] https://t.co/yAqJbFs1OW https://t.co/nuRPJUHJ8m cuTile Python Dec 4"
X Link 2026-01-25T02:30Z [----] followers, [---] engagements
"@t3840 "Wafer-Scale Systems: A Carbon Perspective" Harvard & FAIR at Meta ACM SIGEnergy Energy Informatics Review Aug [--] [----] LLMs CS-3 H100 Wafer-Scale AI Accelerators vs. Single-Chip GPUs Review (33 pp [---] ref) https://x.com/ogawa_tter/status/2011071479715414082 = "Wafer-Scale Systems: A Carbon Perspective" ACM SIGEnergy Energy Informatics Review Aug [--] [----] https://t.co/uIYXOkOtir Cerebras CS-3 vs H100 systems Performance Efficiency & Cost analysis of Wafer-Scale AI Accelerators vs. Single-Chip GPUs Oct [--] https://t.co/yRytINwLQH https://t.co/XpP730P2rN"
X Link 2026-01-25T10:57Z [----] followers, [---] engagements
"= "Cloud-Hardware Co-Design for Memory Bandwidth-Bound HPC Workloads: Performance and Characteristics of Azure HBv5 Virtual Machines" Microsoft & AMD SCA/HPCAsia [----] Jan [--] Custom-designed AMD EPYC 9V64H 8x HBM3 Nov [--] [----] https://x.com/ogawa_tter/status/1859053986198040579 https://dl.acm.org/doi/10.1145/3773656.3773684 = Azure HBv5 Virtual Machines: A Breakthrough in Memory Bandwidth for HPC Nov [--] [----] https://t.co/oMEDMRo4Jt 4x Custom 4th Gen EPYC with HBM3 Memory Bandwidth: [---] TB/s 400-450 GB Up to [---] Zen4 cores SMT disabled [---] Gb/s IB [---] Gb/s per CPU SoC [---] Gb/s: Azure Boost NIC"
X Link 2026-01-26T09:16Z [----] followers, [---] engagements
"Announcing Hortonworks Data Platform (HDP) version [---] GA: Enterprise Hadoop Apr [--] [----] http://t.co/TGG7X2oTOE http://hortonworks.com/blog/announcing-hdp-2-1-general-availability/ http://hortonworks.com/blog/announcing-hdp-2-1-general-availability/"
X Link 2014-04-23T10:39Z [----] followers, [--] engagements
""Common LISP Modules Artificial Intelligence in the Era of Neural Networks and Chaos Theory" M. Watson SAIC [----] http://link.springer.com/book/10.1007/978-1-4612-3186-8 http://link.springer.com/book/10.1007/978-1-4612-3186-8"
X Link 2015-12-27T03:52Z [----] followers, [--] engagements
".@ogawa_tter "The Quantum Software Challenge" What SW Engineers should know about Quantum Computing Oct [--] [----] https://medium.com/@rigetticomputing/the-quantum-software-challenge-1a86eec1ce47 https://medium.com/@rigetticomputing/the-quantum-software-challenge-1a86eec1ce47"
X Link 2016-02-13T14:18Z [----] followers, [--] engagements
".@ogawa_tter Rigetti Computing About 2016/2/13 [--] PhD12 Google Scholar8 arXiv6 MS2 Operations Manager http://rigetti.com/about http://rigetti.com/about"
X Link 2016-02-13T17:49Z [----] followers, [--] engagements
".@ogawa_tter Rigetti Computing Feb [--] [----] "able to accelerate some kinds of machine learning and run highly accurate chemistry simulations""
X Link 2016-02-18T09:02Z [----] followers, [--] engagements
".@ogawa_tter A Practical Quantum Instruction Set Architecture Rigetti Comp Aug [--] [----] https://arxiv.org/abs/1608.03355 https://twitter.com/ogawa_tter/status/764072652214374400 https://arxiv.org/abs/1608.03355 https://twitter.com/ogawa_tter/status/764072652214374400"
X Link 2016-10-17T15:34Z [----] followers, [--] engagements
".@ogawa_tter Emerson W. Pugh Building IBM IBM's [---] and Early [---] Systems Oral History [----] http://ethw.org/Oral-History:Emerson_W._Pugh https://mitpress.mit.edu/authors/emerson-w-pugh http://ethw.org/Oral-History:Emerson_W._Pugh https://mitpress.mit.edu/authors/emerson-w-pugh"
X Link 2017-01-15T02:09Z [----] followers, [--] engagements
".@ogawa_tter IBM System/360 Model [--] Ref P. [--] Jan [----] PDF https://en.wikipedia.org/wiki/IBM_System/360_Model_85 http://bitsavers.trailing-edge.com/pdf/ibm/360/funcChar/A22-6916-1_360-85_funcChar_Jun68.pdf http://www.techrepublic.com/pictures/photos-looking-back-to-the-birth-of-the-mainframe/2/ https://en.wikipedia.org/wiki/IBM_System/360_Model_85 http://bitsavers.trailing-edge.com/pdf/ibm/360/funcChar/A22-6916-1_360-85_funcChar_Jun68.pdf http://www.techrepublic.com/pictures/photos-looking-back-to-the-birth-of-the-mainframe/2/"
X Link 2017-01-15T02:43Z [----] followers, [--] engagements
"= "The DOE Exascale Computing Project (ECP): Overview of Relevant Energy/Power Efforts" EE HPC WG Nov [--] [----] ECP ST Software Stack ST: Runtime System for Application-Level Power Steering Ref: Sep [--] [----] https://x.com/ogawa_tter/status/918089367624630272 https://eehpcwg.llnl.gov/documents/conference/sc17/SC17_workshop_1530_exascale_ang.pdf = Exascale Computing Project Update P. Messina & S. Lee DoE ASCAC Sep [--] [----] PDF https://t.co/CgJxCEzA4G Changes in tech areas https://t.co/y3vj0BvNoD https://x.com/ogawa_tter/status/918089367624630272"
X Link 2017-11-27T09:50Z [----] followers, [--] engagements
"= Oak Ridge Leadership Computing Facility: Recent updates and roadmap forward ASCR BigPanDA SC17 UTA booth Nov [--] [----] PDF Frontier: The OLCF [----] Exascale computer ORNL [----] Summit https://indico.cern.ch/event/680965/contributions/2790231/attachments/1558510/2452141/Wells_UT-ABooth_SC17.pdf https://indico.cern.ch/event/680965/contributions/2790231/attachments/1558510/2452141/Wells_UT-ABooth_SC17.pdf"
X Link 2017-12-11T13:37Z [----] followers, [--] engagements
"= Oak Ridge Leadership Computing Facility: Recent updates and roadmap forward ASCR BigPanDA SC17 UTA booth Nov [--] (NOT 04) [----] Frontier: The OLCF [----] Exascale computer ORNL [----] https://x.com/ogawa_tter/status/940213972027916288 = Oak Ridge Leadership Computing Facility: Recent updates and roadmap forward ASCR BigPanDA SC17 UTA booth Nov [--] [----] PDF https://t.co/bT0lFnNobQ Frontier: The OLCF [----] Exascale computer ORNL [----] Summit https://t.co/Uzx9AbYVtz https://x.com/ogawa_tter/status/940213972027916288 = Oak Ridge Leadership Computing Facility: Recent updates and roadmap forward ASCR"
X Link 2017-12-11T14:24Z [----] followers, [--] engagements
"Rigetti Blake Johnson HPC User Forum Apr [--] [----] PDF Forest An API for quantum computing in the cloud Sep [--] [----] http://www.blakerobertjohnson.com/about https://www.nature.com/news/first-quantum-computers-need-smart-software-1.22590 https://www.rigetti.com/forest https://rigetti.com/ https://www.hpcuserforum.com/presentations/tuscon2018/QCOverview_Rigetti_UFTucson2018.pdf https://www.youtube.com/watchv=VutplvAZxNw http://www.blakerobertjohnson.com/about https://www.nature.com/news/first-quantum-computers-need-smart-software-1.22590 https://www.rigetti.com/forest https://rigetti.com/"
X Link 2018-04-26T18:08Z [----] followers, [--] engagements
"= Quantum Computing Sessions HPC User Forum Apr [--] [----] 2/4 J. Gambetta IBM S. Reinhardt D-Wave PDF Slides D-Wave Qubits Europe users conf Apr 11-12 [----] = https://x.com/ogawa_tter/status/988627912495529984 https://www.hpcuserforum.com/presentations/tuscon2018/QCOverview_DWave_UFTucson2018.pdf https://www.youtube.com/watchv=4s9vEaVJjdA https://www.youtube.com/watchv=TLE6ypLgIo0 = Slides [----] D-Wave Qubits Europe users conference Apr 11-12 https://t.co/afQGcOpPjP D-Wave = Next-gen architecture: Pegasus P6 https://t.co/UqYN55jiTM https://x.com/ogawa_tter/status/988627912495529984"
X Link 2018-04-26T18:44Z [----] followers, [--] engagements
"= Quantum Computing Sessions HPC User Forum Apr [--] [----] 3/4 M. Troyer MSR B. Johnson Rigetti C. Williams NIST Zhu CAS CoE Chris Monroe IonQ = https://ionq.co/ https://x.com/ogawa_tter/status/989554968246927360 https://x.com/ogawa_tter/status/989556678000496640 https://x.com/ogawa_tter/status/989566936034885632 https://x.com/ogawa_tter/status/989561091737899008 "High Performance Quantum Computing" Matthias Troyer Microsoft Research HPC User Forum Apr [--] [----] https://t.co/TCN7t7CAEh PDF https://t.co/DjkWMM1ChQ https://t.co/CJCbp37v1m QC at MSR https://t.co/Ui4nWxROOf Microsoft Quantum"
X Link 2018-04-26T18:45Z [----] followers, [--] engagements
Limited data mode. Full metrics available with subscription: lunarcrush.com/pricing