# ![@NVIDIAAIDev Avatar](https://lunarcrush.com/gi/w:26/cr:twitter::877952584333410305.png) @NVIDIAAIDev NVIDIA AI Developer

NVIDIA AI Developer posts on X about ai, $nvda, inference, cosmos the most. They currently have [------] followers and [---] posts still getting attention that total [------] engagements in the last [--] hours.

### Engagements: [------] [#](/creator/twitter::877952584333410305/interactions)
![Engagements Line Chart](https://lunarcrush.com/gi/w:600/cr:twitter::877952584333410305/c:line/m:interactions.svg)

- [--] Week [-------] +92%
- [--] Month [---------] +49%
- [--] Months [---------] +45%
- [--] Year [---------] +8.40%

### Mentions: [--] [#](/creator/twitter::877952584333410305/posts_active)
![Mentions Line Chart](https://lunarcrush.com/gi/w:600/cr:twitter::877952584333410305/c:line/m:posts_active.svg)

- [--] Month [---] +60%
- [--] Months [---] +40%
- [--] Year [---] +10%

### Followers: [------] [#](/creator/twitter::877952584333410305/followers)
![Followers Line Chart](https://lunarcrush.com/gi/w:600/cr:twitter::877952584333410305/c:line/m:followers.svg)

- [--] Week [------] +1.10%
- [--] Month [------] +2.80%
- [--] Months [------] +14%
- [--] Year [------] +34%

### CreatorRank: [-------] [#](/creator/twitter::877952584333410305/influencer_rank)
![CreatorRank Line Chart](https://lunarcrush.com/gi/w:600/cr:twitter::877952584333410305/c:line/m:influencer_rank.svg)

### Social Influence

**Social category influence**
[stocks](/list/stocks)  #2579 [technology brands](/list/technology-brands)  100% [social networks](/list/social-networks)  1.38% [countries](/list/countries)  1.38% [travel destinations](/list/travel-destinations)  1.38% [finance](/list/finance)  1.38% [cryptocurrencies](/list/cryptocurrencies)  0.69%

**Social topic influence**
[ai](/topic/ai) #2545, [$nvda](/topic/$nvda) #47, [inference](/topic/inference) #155, [cosmos](/topic/cosmos) #164, [model](/topic/model) #1842, [up to](/topic/up-to) #1768, [agentic](/topic/agentic) 5.52%, [generative](/topic/generative) #200, [llm](/topic/llm) 4.14%, [realtime](/topic/realtime) 4.14%

**Top accounts mentioned or mentioned by**
[@nvidia](/creator/undefined) [@huggingface](/creator/undefined) [@awscloud](/creator/undefined) [@grok](/creator/undefined) [@william_r2rclub](/creator/undefined) [@vllmproject](/creator/undefined) [@mistralai](/creator/undefined) [@unslothai](/creator/undefined) [@vllm_project](/creator/undefined) [@microsoft](/creator/undefined) [@primeintellect](/creator/undefined) [@sglproject](/creator/undefined) [@kimmonismus](/creator/undefined) [@stanford](/creator/undefined) [@stepfunai](/creator/undefined) [@pcgeek86](/creator/undefined) [@nvidiaai](/creator/undefined) [@baxate](/creator/undefined) [@googlecloud](/creator/undefined) [@shadeformai](/creator/undefined)

**Top assets mentioned**
[NVIDIA Corp. (NVDA)](/topic/$nvda) [Microsoft Corp. (MSFT)](/topic/microsoft) [IBM (IBM)](/topic/ibm) [ServiceNow Inc (NOW)](/topic/servicenow)
### Top Social Posts
Top posts by engagements in the last [--] hours

"Is your game avatar just not understanding you With the new #AI framework PADL you can more easily command your 3D characters with natural language commands. Learn more about this state-of-the-art paper here: https://nvda.ws/3DuIwFB https://nvda.ws/3DuIwFB"  
[X Link](https://x.com/NVIDIAAIDev/status/1620634601625378817)  2023-02-01T04:06Z 91.8K followers, 138.1K engagements


"We just released [--] million samples of high quality vision language model training dataset for use cases such as: 📄 optical character recognition (OCR) 📊 visual question answering (VQA) 📝 captioning 🤗 Learn more: 📥 Download: https://nvda.ws/4fz2gtB https://nvda.ws/4oyfevu https://nvda.ws/4fz2gtB https://nvda.ws/4oyfevu"  
[X Link](https://x.com/NVIDIAAIDev/status/1955332008890208540)  2025-08-12T18:14Z 91.8K followers, 96.6K engagements


"We just launched an ultra-efficient NVFP4 precision version of Nemotron [--] Nano that delivers up to 4x higher throughput on Blackwell B200. Using our new Quantization Aware Distillation method the NVFP4 version achieves up to 99.4% accuracy of BF16. Nemotron [--] Nano NVFP4: Tech Report: https://nvda.ws/4bj3pp0 https://nvda.ws/4t63z9y https://nvda.ws/4bj3pp0 https://nvda.ws/4t63z9y"  
[X Link](https://x.com/NVIDIAAIDev/status/2016556881712472570)  2026-01-28T17:00Z 91.8K followers, 127.6K engagements


"Kimi K2.5 is now live on GPU-accelerated endpoints for free prototyping so you can quickly start building with a frontier-scale multimodal model instead of just reading about it. Want to get started Weve got you: Step-by-step tutorial Ready-to-run GitHub notebook First inference in minutes not hours 📖 All available in the technical blog https://nvda.ws/4ad6EMw https://nvda.ws/4ad6EMw"  
[X Link](https://x.com/NVIDIAAIDev/status/2019933108317917530)  2026-02-07T00:35Z 91.8K followers, 109.9K engagements


"A new playlist dropped: [--] sessions on NVIDIA Dynamo covering large-scale inference in production including disaggregated serving KV-aware routing MoE and multimodal workloads. Watch on demand Featuring presenters from @basetenco @GoogleAI @haoailab @_llm_d_ @Microsoft @Pinterest @PrimeIntellect @sgl_project @tensormesh and @vllm_project. 🙌 https://www.nvidia.com/en-us/on-demand/playlist/playList-e42aee58-4db9-4ce4-8a6f-c41d8e272d72/ncid=so-twit-451234 https://www.nvidia.com/en-us/on-demand/playlist/playList-e42aee58-4db9-4ce4-8a6f-c41d8e272d72/ncid=so-twit-451234"  
[X Link](https://x.com/NVIDIAAIDev/status/2021615212315484376)  2026-02-11T16:00Z 91.8K followers, [----] engagements


"🧑🏻🍳 Introducing the NVIDIA Cosmos Cookoff a virtual four-week physical AI challenge for robotics AV and vision AI builders sponsored by @nebiusai and @milestonesys. Build with NVIDIA Cosmos Reason [--] and Cosmos Cookbook recipesfrom egocentric robot reasoning to physical plausibility checks and trafficaware models. 🗓 Jan [--] Feb [--] 👥 Solo or teams (up to 4) 🏆 Prizes include $5000 an NVIDIA DGX Spark and more 🧑⚖ With judges from @DatatureAI @huggingface @nebiusai @getnexar and @nvidia Sign up and show whats possible with physical AI ➡ https://nvda.ws/3Z9lrCO https://nvda.ws/3Z9lrCO"  
[X Link](https://x.com/NVIDIAAIDev/status/2016919273780674803)  2026-01-29T17:00Z 91.8K followers, 172.3K engagements


"👀 Discover the three ways NVFP4 transforms AI training and inference delivering more performance and efficiency and enabled by extreme codesign. 👉 https://nvda.ws/4r9IHx2 https://nvda.ws/4r9IHx2"  
[X Link](https://x.com/NVIDIAAIDev/status/2019805789242753463)  2026-02-06T16:10Z 91.8K followers, 14.5K engagements


"ICYMI - @liu_mingyu Vice President of Research at NVIDIA and @kimmonismus from @getsuperintel recently chatted about how NVIDIA Cosmos is reshaping AI for robots AVs and vision-driven systems. Watch the full livestream ➡ Or catch Ming-Yu at #NVIDIAGTC ➡ https://nvda.ws/4azcfwN https://nvda.ws/3ZSTJuB https://nvda.ws/4azcfwN https://nvda.ws/3ZSTJuB"  
[X Link](https://x.com/NVIDIAAIDev/status/2022037999395991744)  2026-02-12T20:00Z 91.8K followers, [----] engagements


"30M downloads and counting for the NVIDIA Nemotron family on @huggingface 🤗 We're grateful for the incredible community that has made this possible. Get started with Nemotron: https://nvda.ws/4q8MtVP https://nvda.ws/4q8MtVP"  
[X Link](https://x.com/NVIDIAAIDev/status/2018731215440760957)  2026-02-03T17:00Z 91.8K followers, 18.5K engagements


"Explore where physical AI is headed next with the people shaping it. On Feb. [--] at [--] a.m. PT join MingYu Vice President of Research at NVIDIA and @kimmonismus from Superintelligence for a conversation on how NVIDIA Cosmos reframes AI for a world of robots AVs and vision-driven systems. We'll dig into: ✅ Cosmos' origins and designing models for the physical world ✅ The scientific ideas and architectures behind Cosmos and its open models ✅ How open reasoning-capable physical AI is changing how teams prototype deploy and iterate ✅ What a future of AI systems that understand space time and"  
[X Link](https://x.com/NVIDIAAIDev/status/2020939396640538690)  2026-02-09T19:14Z 91.8K followers, 25.2K engagements


"Did you know theres an NVIDIA DGX Spark powering AI research at the South Pole ❄ At the @UWMadisons IceCube Neutrino Observatory in Antarctica researchers are using DGX Spark to run AI models that help study the universes most cataclysmic eventsthrough subatomic particles called neutrinos. The NVIDIA DGX Spark desktop supercomputer brings datacenterclass AI performance to lab benches faculty offices and student systems worldwide. Discover how DGX Spark fuels groundbreaking AI research at universities including @ASU @Harvard ISTA @msstate @nyuniversity @Stanford and @UDelaware. 🔗 📸 DGX Spark"  
[X Link](https://x.com/NVIDIAAIDev/status/2021993454805381183)  2026-02-12T17:03Z 91.8K followers, 19.4K engagements


"✨ Thank you to our incredible community of developers and researchers driving AI innovation one model one commit and one breakthrough at a time. 🤗 Were thrilled to celebrate 50000+ followers on @HuggingFace. Keep building with us: https://huggingface.co/nvidia https://huggingface.co/nvidia"  
[X Link](https://x.com/NVIDIAAIDev/status/2022023402047799389)  2026-02-12T19:02Z 91.8K followers, 35.7K engagements


"👨🏫📝 Apply to the NVIDIA DLI Teaching Kit Program to explore the updated Generative AI Teaching Kit. It includes new lecture videos for flipped or selfpaced learning teaching materials aligned with the NVIDIA GenAI Associate certification free educator exams and limited 50% student discounts available through May [--] [----]. ➡ Created in collaboration with the Thayer School of Engineering at @dartmouth. https://developer.nvidia.com/blog/nvidia-deep-learning-institute-releases-new-generative-ai-teaching-kit/"  
[X Link](https://x.com/NVIDIAAIDev/status/2022369327958364216)  2026-02-13T17:56Z 91.8K followers, [----] engagements


"From interactive hair simulation to an #AI system that can auto-generate realistic 3D character motion by watching 2D videos check out the [--] papers from #NVIDIAResearch accepted into #SIGGRAPH2023. Read the blog with links to each paper: https://nvda.ws/3nniRKc https://nvda.ws/3nniRKc"  
[X Link](https://x.com/NVIDIAAIDev/status/1653453776081018880)  2023-05-02T17:37Z 91.7K followers, 463.1K engagements


"🚨 GeForce RTX 5090s are available at the Gear Store in the park right now at #GTC25 [--] units are available for the next [--] minutes with more coming tomorrow. Come say hi"  
[X Link](https://x.com/NVIDIAAIDev/status/1902454685153554438)  2025-03-19T20:18Z 91.8K followers, 68.1K engagements


"NVIDIAs Graduate Fellowship Program is now accepting applications for the [--------] academic year. Selected Ph.D. students receive tuition and stipend coverage up to $60K plus mentorship and technical support from top NVIDIA researchers during an NVIDIA internship. If youre advancing work in AI robotics computer graphics autonomous vehicles healthcare HPC or related fields this is your moment. 📅 Apply by Sept. [--] 2025: https://nvda.ws/3UsbEpe https://nvda.ws/3UsbEpe"  
[X Link](https://x.com/NVIDIAAIDev/status/1955690818523591087)  2025-08-13T18:00Z 91.5K followers, 112.2K engagements


"🎉 Congratulations @IBM on the launch of your new Granite [---] family of open LLMs. 🙌 Granite 4.0s design for maximum efficiency trust and security is a perfect match for secure production-ready deployment on any NVIDIA accelerated infrastructure with NIM. Get started with Granite [---] -- now available on NVIDIA NIM ➡ https://build.nvidia.com/nvidia/bring-llm-to-nim The next generation of open-source AI is here. Meet IBM Granite [---]. We've engineered these open performant and trusted small language models to solve key enterprise challenges. Discover how Granite [---] delivers exceptional"  
[X Link](https://x.com/NVIDIAAIDev/status/1975277964687515866)  2025-10-06T19:12Z 91.8K followers, 29.5K engagements


"The largest advancement of the CUDA platform since its creation in [----] is here 👀 Introducing CUDA Tile a tile-based programming model that provides the ability to write algorithms at a higher level and abstract away the details of specialized hardware such as tensor cores. Read the technical blog 👉 https://developer.nvidia.com/blog/focus-on-your-algorithm-nvidia-cuda-tile-handles-the-hardware https://developer.nvidia.com/blog/focus-on-your-algorithm-nvidia-cuda-tile-handles-the-hardware"  
[X Link](https://x.com/NVIDIAAIDev/status/1996976702732620271)  2025-12-05T16:15Z 91.3K followers, 243.6K engagements


"🎂 One year of Cosmos and what a journey it's been. 🌑 We're celebrating with some incredible milestones: 🚀 5M total downloads across the Cosmos ecosystem 🧠 Cosmos Reason is the #1 model on the physical reasoning leaderboard with 2M+ downloads on @HuggingFace 🔮 Cosmos Predict is the #1 open model on the physical AI generation leaderboard with 2M+ downloads Thank you to our amazing developer community for making this possible. Here's to pushing the boundaries of world foundation models together 📚 Explore Models & Datasets: 🧑🏻🍳 Read the Cosmos Cookbook: https://nvda.ws/4qevli8"  
[X Link](https://x.com/NVIDIAAIDev/status/2008976940963655853)  2026-01-07T19:00Z 91.2K followers, 28.8K engagements


"✨ One year after the debut of "Project Digits" DGX Spark has transformed from concept to reality and is now available from NVIDIA and OEM partners. See it in action lighting up the show floor at #CES2026. #SparkSomethingBig"  
[X Link](https://x.com/NVIDIAAIDev/status/2011876035944485085)  2026-01-15T19:00Z 91.6K followers, 37.8K engagements


"Looking to add reasoning to your vision AI AV or robotics pipeline On Feb. [--] at [--] a.m. PT join NVIDIA and @intbotai engineers as they walk through four NVIDIA Cosmos Cookbook recipes including: ✅ Robothuman interaction with egocentric social + physical reasoning ✅ Intelligent transportation with posttraining Cosmos Reason ✅ Rilter unreliable synthetic video with physical plausibility prediction ✅ Search summarize and act on live or recorded streams with video analytics AI agent Tune in and turn these recipes into your next project ➡ https://nvda.ws/4qVYXS5 https://nvda.ws/4qVYXS5"  
[X Link](https://x.com/NVIDIAAIDev/status/2018399017005203893)  2026-02-02T19:00Z 91.2K followers, [----] engagements


"Nemotron just landed on the ViDoRe V3 leaderboard testing RAG capabilities across visuals documents and reasoning tasks. Nemotron-colembed-8B-V2 pushes ahead with sharper visual comprehension for complex charts better multilingual reading and improved accuracy through model merging. Model: HF Blog: https://nvda.ws/4bwBHoQ https://nvda.ws/3O8NVu5 https://nvda.ws/4bwBHoQ https://nvda.ws/3O8NVu5"  
[X Link](https://x.com/NVIDIAAIDev/status/2019072889660625010)  2026-02-04T15:37Z 91.2K followers, [----] engagements


"New model just dropped 👀 Its a sparse MoE model from @StepFun_ai featuring up to a 1M context window built for fast reasoning and agent-style workflows. Get started at and test it out today no setup no friction. http://build.nvidia.com/stepfun-ai/step-3.5-flash http://build.nvidia.com/stepfun-ai/step-3.5-flash"  
[X Link](https://x.com/NVIDIAAIDev/status/2019169094281351498)  2026-02-04T22:00Z 91.7K followers, 12.5K engagements


"👀 @vllm_project is raising the bar for largescale inference on NVIDIA Blackwell using NVFP4 to hit 26.2K tokens/second per GPU for prefill and 10.1K tokens/second per GPU for decode on GB200 NVL72 for DeepSeekstyle MoE models including DeepSeek R1/V3/V3.1. Working closely with the OSS community @AIatMeta and our engineers vLLM optimized the engine for these models delivering major throughput and efficiency gains over Hopper.⚡ These gains are driven by: ✅Lower-precision operations ✅Kernel fusion ✅Scaling down prefill via weight offloading ✅Minimized chunking overheads Dive into the details"  
[X Link](https://x.com/NVIDIAAIDev/status/2019187101536043330)  2026-02-04T23:11Z 91.3K followers, 30.2K engagements


"Build your AI agent to "comprehend" complex PDFs nested tables and charts with superior precision 📊 We just dropped a full guide on building a high-throughput Intelligent Document Processing pipeline with Nemotron RAG. Check it out ➡ https://nvda.ws/3ZQ8y0C https://nvda.ws/3ZQ8y0C"  
[X Link](https://x.com/NVIDIAAIDev/status/2019456089054413036)  2026-02-05T17:00Z 91.8K followers, [----] engagements


"Build Reasoning Agents For Physical AI Cosmos Labs https://x.com/i/broadcasts/1ZkKzZRkqaqKv https://x.com/i/broadcasts/1ZkKzZRkqaqKv"  
[X Link](https://x.com/NVIDIAAIDev/status/2019456866972234153)  2026-02-05T17:03Z 91.2K followers, [----] engagements


"✨ Got questions about DGX Spark / GB10 Join us live to get practical answers from engineers who work with DGX Spark every day. Well cover key topics like performance benefits use cases and integration approaches. Whether youre exploring how Spark can support your development workflows or looking to get more out of your current setup this livestream is for you. 📆 Friday February [--] at [--] a.m. Pacific 🎥 Watch here on X or on YouTube: https://youtube.com/live/nRo-tQC-mEY https://youtube.com/live/nRo-tQC-mEY"  
[X Link](https://x.com/NVIDIAAIDev/status/2019486181155610926)  2026-02-05T19:00Z 91.5K followers, [----] engagements


".@JusttFintech cut their error rate by 25% using NVIDIA Nemotron Parse⚡ See how they built a high-fidelity GPU-accelerated extraction pipeline for chargeback evidence. Read the full case study: https://developer.nvidia.com/case-studies/justt-scaled-chargeback-document-extraction-using-nvidia-nemotron-parsencid=so-twit-720777 https://developer.nvidia.com/case-studies/justt-scaled-chargeback-document-extraction-using-nvidia-nemotron-parsencid=so-twit-720777"  
[X Link](https://x.com/NVIDIAAIDev/status/2019516380056309833)  2026-02-05T21:00Z 91.2K followers, [----] engagements


"🤖 Build and deploy agentic applications with us and @Microsoft. At our AI Apps & Agents Dev Days youll learn how to overcome real-world challenges and experiment with whats next in AI-driven apps and agent design. 📆 February 10: Build a Voice-Enabled AI Agent in Minutes ➡ 📆 March 11: Coordinating Secure GPU-Accelerated Agents with Foundry Control Plane on Azure ➡ 💻 On-demand: Build a Deep Research Agent for Enterprise Data with Scalable Performance ➡ Join the morning (9-10:00am PT) or night session (6-7:00pm PT). 🔗 Learn more about the technical series: https://nvda.ws/3ZQJDKm"  
[X Link](https://x.com/NVIDIAAIDev/status/2019531479307415616)  2026-02-05T22:00Z 91.2K followers, [----] engagements


"@pcgeek86 @NVIDIAAI @nvidia Hi Trevor - The Nemotron forums would be best https://forums.developer.nvidia.com/c/ai-data-science/nvidia-nemotron https://forums.developer.nvidia.com/c/ai-data-science/nvidia-nemotron"  
[X Link](https://x.com/NVIDIAAIDev/status/2019544382966505708)  2026-02-05T22:51Z 91.2K followers, [---] engagements


"Inference Office Hours with SGLang: Performance Optimizations for LLM Serving https://x.com/i/broadcasts/1nAJEEpZyBAJL https://x.com/i/broadcasts/1nAJEEpZyBAJL"  
[X Link](https://x.com/NVIDIAAIDev/status/2019584208193352175)  2026-02-06T01:29Z 91.4K followers, [----] engagements


"DGX Spark Live: Your Questions Answered Vol. [--] https://x.com/i/broadcasts/1eaKbjaryMVKX https://x.com/i/broadcasts/1eaKbjaryMVKX"  
[X Link](https://x.com/NVIDIAAIDev/status/2019848547827626479)  2026-02-06T18:59Z 91.7K followers, [----] engagements


"How much memory would you win back if every KV cache were [----] smaller KVTC a new #NVIDIAResearch method compresses LLM KV caches for storage like a media codec. If your serving stack is KV-bound this is a new primitive for multi-turn and long-context systems. 📗 Paper: https://arxiv.org/pdf/2511.01815.pdf https://arxiv.org/pdf/2511.01815.pdf"  
[X Link](https://x.com/NVIDIAAIDev/status/2019868689907216876)  2026-02-06T20:19Z 91.7K followers, 12.7K engagements


"Want a chance to attend #NVIDIAGTC Just one week left to enter: We're partnering with our GTC community to give away Golden Tickets 🎫 including: ✅GTC Conference pass ✅VIP seating at NVIDIA CEO Jensen Huangs keynote ✅NVIDIA DGX Spark ✅Exclusive Happy Hour at NVIDIA Headquarters ✅GTC Training Lab pass ✅NVIDIA merchandise Follow these accounts to be the first to know each giveaway's unique entry requirements 🔽 NVIDIA team: @ctnzr @Baxate / @NaderLikeLadder Partners: @ollama @huggingface / @pollenrobotics Get contest details: Terms & conditions apply. https://nvda.ws/4rdhVUk"  
[X Link](https://x.com/NVIDIAAIDev/status/2020972482253979807)  2026-02-09T21:26Z 91.7K followers, [----] engagements


"🙌 Explore the next generation of AI at #NVIDIAGTC kicking off with a keynote from our CEO Jensen Huang followed by expert-led sessions hands-on demos hackathons exclusive networking opportunities and more. Don't miss the premier AI conference. 📆 March [----] in San Jose California. Register now before the event is sold out 👉 https://nvda.ws/4tqi7kz https://nvda.ws/4tqi7kz"  
[X Link](https://x.com/NVIDIAAIDev/status/2020988582396096815)  2026-02-09T22:30Z 91.7K followers, [----] engagements


"How To Adapt AI for Low-Resource Languages with NVIDIA Nemotron https://x.com/i/broadcasts/1lPKqvMwlRZGb https://x.com/i/broadcasts/1lPKqvMwlRZGb"  
[X Link](https://x.com/NVIDIAAIDev/status/2021207570061590579)  2026-02-10T13:00Z 91.7K followers, [----] engagements


"👀Global hackathon with @MistralAI 👇 Introducing Mistral AI's biggest hackathon ever 📅 Feb [--] - Mar [--] 🌍 Paris London NY SF Tokyo Singapore Sydney & online [--] hours. The best hackers. 🤝 Partners: @wandb @nvidia @awscloud @HackIterate 🏆 $200K in prizes. Special awards from @elevenlabsio @huggingface https://t.co/rkqDc0pUy8 Introducing Mistral AI's biggest hackathon ever 📅 Feb [--] - Mar [--] 🌍 Paris London NY SF Tokyo Singapore Sydney & online [--] hours. The best hackers. 🤝 Partners: @wandb @nvidia @awscloud @HackIterate 🏆 $200K in prizes. Special awards from @elevenlabsio @huggingface"  
[X Link](https://x.com/NVIDIAAIDev/status/2021305668901195796)  2026-02-10T19:30Z 91.7K followers, [----] engagements


"@willccbb @huggingface @Alibaba_Qwen @allen_ai @arcee_ai @Meta @nvidia @Zai_org @PrimeIntellect 🙌"  
[X Link](https://x.com/NVIDIAAIDev/status/2021630250740810157)  2026-02-11T16:59Z 91.7K followers, [---] engagements


"👀 New livestream this Friday. Join our discussion with @Stanford researchers for an exciting look at Biomni a next-generation biomedical AI agent transforming how labs explore data and design experiments. Discover how Biomni powered by NVIDIA DGX Spark turns complex biomedical workloads into fast automated insights giving teams a powerful launchpad for rapid experimentation evaluation and seamless scaling. 📆 Friday February [--] at [--] a.m. Pacific 🎥 Watch here on X or on YouTube: https://www.youtube.com/watchv=8v_dfwpAzdY https://www.youtube.com/watchv=8v_dfwpAzdY"  
[X Link](https://x.com/NVIDIAAIDev/status/2021675609374536087)  2026-02-11T20:00Z 91.7K followers, [----] engagements


"Ready to build an AI agent Our new guide covers everything from agent architecture to state management with LangGraph all powered by NVIDIA Nemotron on @OpenRouterAI. Learn advanced concepts and get started instantly with a one-click NVIDIA Launchable. 🔗 https://nvda.ws/468NmHt https://nvda.ws/468NmHt"  
[X Link](https://x.com/NVIDIAAIDev/status/1968087506550702271)  2025-09-16T23:00Z 91.8K followers, 175.8K engagements


"Cosmos is coming to #NVIDIAGTC. 👋 From deep dives to handson demos dont miss our sessions on scaling generative world models simulation and nextgen AI systems. Add session to your schedule now 🔗 https://nvda.ws/4rjM3NR https://nvda.ws/4rjM3NR"  
[X Link](https://x.com/NVIDIAAIDev/status/2021328320705462743)  2026-02-10T21:00Z 91.8K followers, 12.9K engagements


"Securely experiment and build your own specialized agents as the 671-billion-parameter DeepSeek-R1 model is now available as an NVIDIA NIM microservice in preview on Learn more ➡ https://nvda.ws/4grQaBq http://build.nvidia.com https://nvda.ws/4grQaBq http://build.nvidia.com"  
[X Link](https://x.com/NVIDIAAIDev/status/1885110101079642576)  2025-01-30T23:37Z 91.8K followers, 644.7K engagements


"This is an incredible performance breakthrough from @UnslothAI. 12x faster fine-tuning 35% less VRAM all with no loss in accuracy enables fine-tuning of MoE models like gpt-oss-20b on just [--] GB of VRAM. You can now train MoE models [--] faster with 35% less VRAM via our new Triton kernels (no accuracy loss). Train gpt-oss locally on 12.8GB VRAM. In collab with @HuggingFace Unsloth trains DeepSeek Qwen3 GLM faster. Repo: https://t.co/aZWYAtakBP Blog: https://t.co/3wRiBxVJB6 https://t.co/MZke9gtISU You can now train MoE models [--] faster with 35% less VRAM via our new Triton kernels (no accuracy"  
[X Link](https://x.com/NVIDIAAIDev/status/2021398069523382737)  2026-02-11T01:37Z 91.8K followers, 127.4K engagements


"From PyTorch to production-ready engines in days not weeks 💡 With TensorRT LLM's new AutoDeploy feature developers can automate the heavy lifting of inference optimization and reduce deployment time. Dive into architecture performance numbers and examples 👉 https://nvda.ws/4qwFEhe https://nvda.ws/4qwFEhe"  
[X Link](https://x.com/NVIDIAAIDev/status/2021720909548056773)  2026-02-11T23:00Z 91.8K followers, [----] engagements


"Most 3DGS segmentation tools either pretrain per scene or lock errors into a feature field you cant undo. ArtisanGS instead turns a few 2D masks into editable 3D object selections via Cutie tracking + blackbox splat aggregation then lets you iteratively correct mistakes with consistent 2D/3D selection modes. 📗 #NVIDIAResearch paper: https://arxiv.org/abs/2602.10173 https://arxiv.org/abs/2602.10173"  
[X Link](https://x.com/NVIDIAAIDev/status/2022400382991835282)  2026-02-13T20:00Z 91.8K followers, 35.5K engagements


"🎉 Congrats to @MistralAI on the launch of the new Mistral [--] open family of models Mistral Large [--] delivers up to [--] performance on NVIDIA NVL72 by stacking the benefits of large scale expert parallelism inference disaggregation and accuracy-preserving NVFP4 low-precision inference. Trained on NVIDIA Hopper these models deliver a range of size scale and efficiency to meet a wide variety of needs: ✅ Mistral Large 3: multimodal and multilingual MoE model with a total parameter count of 675B ✅ Ministral 3: small dense models each with base instruct and reasoning variants Learn more ➡"  
[X Link](https://x.com/NVIDIAAIDev/status/1995920443518095599)  2025-12-02T18:18Z 86.7K followers, [----] engagements


"Portfolio optimization often faces a trade-off between speed and complexity. The new Quantitative Portfolio Optimization developer example accelerates backtesting and analysis with high-performance computing. Read more 👉 https://nvda.ws/48bMmSr https://nvda.ws/48bMmSr"  
[X Link](https://x.com/NVIDIAAIDev/status/1995948592805097705)  2025-12-02T20:10Z 86.2K followers, [----] engagements


"We kicked off #NeurIPS2025 with the Nemotron Summit bringing researchers and developers together to explore how open models are reshaping AI. From deep dives into open models datasets and techniques to a standout session with @ServiceNow every talk was packed with insights and ideas. Thanks to everyone who joined us see you next time 🙌"  
[X Link](https://x.com/NVIDIAAIDev/status/1995996115607425441)  2025-12-02T23:18Z 87K followers, [----] engagements


"Building physical AI for autonomous vehicles robotics or smart factories You know the pain of training datanot enough expensive and doesnt scale. Check out our deep dive on deploying NVIDIA Cosmos WFMs on @awscloud to generate synthetic data at scale. Learn how to: ✔ Generate synthetic data with real-time inference and batch inference ✔ Enable reasoning for physical AI systems Read the full guide: https://nvda.ws/4ozp9jx https://nvda.ws/4ozp9jx"  
[X Link](https://x.com/NVIDIAAIDev/status/1996006468684755320)  2025-12-03T00:00Z 86.7K followers, [----] engagements


"Day [--] of #NeurIPS2025 ✅ Thanks to everyone who stopped by the booth watched the livestream and joined us at one of our talks. See you tomorrow 💚 Side note: Where's the best burrito in San Diego"  
[X Link](https://x.com/NVIDIAAIDev/status/1996036669112205567)  2025-12-03T02:00Z 86.8K followers, [----] engagements


"Todays Cosmos Cookbook Special 🍽 @Voxel51's recipe that uses Cosmos Transfer [---] to address scarcity in biodiverse data by taking a few existing moth images and generating rich field-like agricultural scenes. 📖 Transform lab-style BioTrove moth shots into photorealistic outdoor scenarios while preserving structure and identity. 📖 Scale up data diversity for rare underrepresented species and close the labtofield domain gap. 📖 Build an endtoend pipeline with Python-only inference edge-based control and FiftyOne visualization and search. Read the full recipe 📖 Download Cosmos Transfer 2.5"  
[X Link](https://x.com/NVIDIAAIDev/status/1996051766916362741)  2025-12-03T03:00Z 86.8K followers, [----] engagements


"Learn how @EPRINews is using NVIDIA DGX Spark to push the boundaries of AI innovation. Our R&D team uses DGX Spark to run advanced open-weight models including get-oss-120b and Qwen-3-Embedding-8Bpowering a secure AI-driven document research assistant. This compact yet powerful system enables small teams to achieve high-quality results in the enterprise infrastructure making custom AI workflows more accessible and compliant. Matt Corbett Director Information Technology Enterprise Infrastructure & Networking EPRI 🎥 Catch the full conversation: #SparkSomethingBig ✨"  
[X Link](https://x.com/NVIDIAAIDev/status/1996278261425308098)  2025-12-03T18:00Z 86.4K followers, [----] engagements


"Todays Cosmos Cookbook Special 🍽 A recipe that lets you generate and build realistic physically accurate and rule-aware sports videos with smooth player motions using LoRA post-trained on Cosmos Predict [---]. 📖 Create videos that capture the natural kinematics of athletes across various sports 📖 Consistently display jerseys teams and identities across the whole play 📖 Generate broadcastquality clips for analytics simulation or fanfacing content Read the full recipe 📖 Download Cosmos Predict [---] 👉 https://nvda.ws/4rx25EN https://nvda.ws/3MBqRTU https://nvda.ws/4rx25EN"  
[X Link](https://x.com/NVIDIAAIDev/status/1996323595614556315)  2025-12-03T21:00Z 86.7K followers, [----] engagements


"Todays Cosmos Cookbook Special 🍽 A recipe that uses an endtoend workflow to generate photorealistic synthetic data for traffic scenarios. 📖 Simulate rich multiagent traffic scenes in a virtual city. 📖 Turn simulation into highfidelity urban video data. 📖 Produce scalable datasets for perceptive VLMs in smart city applications. Read the full recipe 📖 Download Cosmos 👉 https://nvda.ws/48uDHMB https://nvda.ws/4plnNtS https://nvda.ws/48uDHMB https://nvda.ws/4plnNtS"  
[X Link](https://x.com/NVIDIAAIDev/status/1996731352201179484)  2025-12-05T00:00Z 86.6K followers, [----] engagements


"Kernel Challenge #2 is LIVE In partnership with @GPU_MODE we've dropped the next high-performance task: NVFP4 GEMM. Prove your kernel optimization skills and push the limits of low-precision computing. 📝 Problem: NVFP4 GEMM 🗓 Deadline: Dec [--] 🔗 Join Now: https://nvda.ws/3Y7qRxs https://nvda.ws/3Y7qRxs"  
[X Link](https://x.com/NVIDIAAIDev/status/1996746344510153193)  2025-12-05T01:00Z 86.8K followers, [----] engagements


"What would you do with unlimited compute ⚡🦾 We asked NVIDIA Research Directors @ArashVahdat and Pavlo Molchanov that question at #NeurIPS2025 and their answers might surprise you. Get a glimpse into whats coming for AI in [----] by watching the interview highlights"  
[X Link](https://x.com/NVIDIAAIDev/status/1996762460636774817)  2025-12-05T02:04Z 86.7K followers, [----] engagements


"CUDA Tile is the future of GPU programming and it's in Python. 🐍 cuTile Python in CUDA [----] lets you focus on your algorithm not Tensor Core specifics. Simplify your HPC/AI code now. ✨ 🔗 https://nvda.ws/44LtUPJ https://nvda.ws/44LtUPJ"  
[X Link](https://x.com/NVIDIAAIDev/status/1996987939599847650)  2025-12-05T17:00Z 87K followers, 31.6K engagements


"NVIDIA CUDA Toolkit [----] just landed bringing the largest update in [--] years. Read 🔗 https://developer.nvidia.com/blog/nvidia-cuda-13-1-powers-next-gen-gpu-programming-with-nvidia-cuda-tile-and-performance-gainsncid=so-twit-259531 https://developer.nvidia.com/blog/nvidia-cuda-13-1-powers-next-gen-gpu-programming-with-nvidia-cuda-tile-and-performance-gainsncid=so-twit-259531"  
[X Link](https://x.com/NVIDIAAIDev/status/1997003254321090677)  2025-12-05T18:00Z 86.7K followers, [----] engagements


"NVIDIA CUDA Toolkit [----] just landed bringing the largest update in [--] years. Read 🔗 https://developer.nvidia.com/blog/nvidia-cuda-13-1-powers-next-gen-gpu-programming-with-nvidia-cuda-tile-and-performance-gainsncid=so-twit-259531 https://developer.nvidia.com/blog/nvidia-cuda-13-1-powers-next-gen-gpu-programming-with-nvidia-cuda-tile-and-performance-gainsncid=so-twit-259531"  
[X Link](https://x.com/NVIDIAAIDev/status/1997003446726414343)  2025-12-05T18:01Z 86.7K followers, [----] engagements


"Proud of our NVIDIA Grandmasters Ivan Sorokin and Jean-Franois Puget who took 1st place in the Kaggle ARC Prize [----] 👏 Their extended solution also resulted in a notable ARC-AGI-2 leaderboard score of 29.72% at a super economical 20cents/task using synthetic data test-time training and NVIDIA NeMo tools a meaningful step toward more efficient AI reasoning. Hear the behind-the-scenes look at how they did it 👇"  
[X Link](https://x.com/NVIDIAAIDev/status/1997004131974717804)  2025-12-05T18:04Z 86.6K followers, [----] engagements


"Congratulations to our NVIDIA Kaggle Grandmasters on their winning solution in the ARC Prize [----] competition 🎉 The team fine-tuned a compact model that out-reasoned massive systems proving synthetic data + adaptive reinforcement learning can punch far above their weight. 🔗 Details: https://developer.nvidia.com/blog/nvidia-kaggle-grandmasters-win-artificial-general-intelligence-competition/ncid=so-twit-156197 Announcing the ARC Prize [----] Top Score & Paper Award winners The Grand Prize remains unclaimed Our analysis on AGI progress marking [----] the year of the refinement loop"  
[X Link](https://x.com/NVIDIAAIDev/status/1997012155955400863)  2025-12-05T18:36Z 86.7K followers, [----] engagements


"DGX Spark Live: Backend Development with Local LLM Inference https://x.com/i/broadcasts/1BRJjgedapdxw https://x.com/i/broadcasts/1BRJjgedapdxw"  
[X Link](https://x.com/NVIDIAAIDev/status/1997018237134164476)  2025-12-05T19:00Z 86.7K followers, [----] engagements


"Todays Cosmos Cookbook Special 🍽 A recipe to post-train Cosmos Reason into a physics-savvy critic that judges whether generated videos obey real-world physics. 📖 Score videos for physical plausibility 📖 Detect physically inaccurate issues like impossible trajectories or bad collisions 📖 Incorporate physicsaware rewards into your generation or RL loops to keep the models grounded Read the full recipe 📖 https://nvda.ws/4pTI0GZ https://nvda.ws/4pTI0GZ"  
[X Link](https://x.com/NVIDIAAIDev/status/1997034449213506002)  2025-12-05T20:04Z 86.8K followers, [----] engagements


"Accelerate your research and transform the classroom with AI 📚 We've got a full lineup of Higher Education and Research session replays now available on demand. Start watching the free playlist here 👇 https://nvda.ws/3KmMg2J https://nvda.ws/3KmMg2J"  
[X Link](https://x.com/NVIDIAAIDev/status/1998106505183138080)  2025-12-08T19:04Z 86.8K followers, [----] engagements


"📊 New learning pathway - Speed Up Data Analytics on GPUs Built in partnership with @googlecloud get hands-on with NVIDIA cuDF to accelerate pandas code with no code changes. 🔗 https://nvda.ws/3KDEth1 https://nvda.ws/3KDEth1"  
[X Link](https://x.com/NVIDIAAIDev/status/1998120396910895314)  2025-12-08T20:00Z 86.8K followers, [----] engagements


"AI is revolutionizing technology and at the core is agentic AI. Get a comprehensive introduction to the new self-paced course focusing on autonomous software agents that perceive reason act and learn within dynamic environments. In this course you'll: ✅ Understand the core building blocks: perception reasoning action and feedback ✅ Explore real-world applications across industries ✅ Learn how to start building a simple agent yourself Enroll today 👉 https://nvda.ws/48rJTVy https://nvda.ws/48rJTVy"  
[X Link](https://x.com/NVIDIAAIDev/status/1998165695465463898)  2025-12-08T23:00Z 86.8K followers, [----] engagements


"Another open source model drop 🎁 Congrats @MistralAI on releasing [--] new open agentic coding models. Experiment with the Devstral [--] (123B) model on the NVIDIA API catalog and try it out on the NVIDIA AI stack 👉 https://nvda.ws/4iJlyOh https://nvda.ws/4iJlyOh"  
[X Link](https://x.com/NVIDIAAIDev/status/1998418962355253509)  2025-12-09T15:46Z 87.6K followers, 10.4K engagements


"🧵 What it takes to run AI inference at massive scale featuring @awscloud At #AWSreInvent the Amazon Ads team shared a behind-the-scenes look at running NVIDIA Dynamo at production scale for high-performance AI inference. Key takeaways 👇"  
[X Link](https://x.com/NVIDIAAIDev/status/1998543183320375559)  2025-12-10T00:00Z 86.7K followers, [----] engagements


"@awscloud [--] The measured results With this setup they reported: 50% higher throughput from disaggregated inference 2040% lower latency from KV-cacheaware routing Measured on real production workloads"  
[X Link](https://x.com/NVIDIAAIDev/status/1998543195467071592)  2025-12-10T00:00Z 86.7K followers, [----] engagements


"@awscloud [--] Takeaway At large scale inference performance is a systems problem. This session showed how targeted architectural choices not just faster hardware can materially change throughput and latency 🔗 https://nvda.ws/4ppZ5s5 https://nvda.ws/4ppZ5s5"  
[X Link](https://x.com/NVIDIAAIDev/status/1998543197723578557)  2025-12-10T00:00Z 86.7K followers, [----] engagements


"Looking to develop your skills 👀 Join our hands-on labs at #SIGGRAPHAsia2025 to get trained in AI robotics OpenUSD and more. 📆 Register today: https://nvda.ws/491aTv9 https://nvda.ws/491aTv9"  
[X Link](https://x.com/NVIDIAAIDev/status/1998558282646999467)  2025-12-10T01:00Z 86.8K followers, [----] engagements


"Dont forget to submit your research proposal for the #NVIDIAAcademicGrant Program by December [--] 📅 We invite you to submit proposals in: 💡 Generative AI Alignment & Inference ⚡ Generative AI Training and Model Development 🤖 Robotics and Edge AI Get access to cutting-edge resources: NVIDIA DGX Spark RTX PRO [----] AGX Developer Kits and A100 GPU hours. Apply now: https://nvda.ws/4iMsg64 https://nvda.ws/4iMsg64"  
[X Link](https://x.com/NVIDIAAIDev/status/1998784775197171932)  2025-12-10T16:00Z 87.2K followers, [----] engagements


"The path to scalable quantum computing (QC) is being supercharged by AI. 💡 Dive into our Nature Communications review to see how state-of-the-art AI is accelerating everything from qubit design and control to real-time error correction paving the way for useful fault-tolerant QC 👉 https://nvda.ws/4oTt0Id https://nvda.ws/4oTt0Id"  
[X Link](https://x.com/NVIDIAAIDev/status/1998890473260245238)  2025-12-10T23:00Z 87.3K followers, [----] engagements


"Spark Curiosity. Empower Creation. 🎁 The best gifts inspire action. DGX Spark turns curiosity into creation offering powerful energy-efficient AI compute for learners developers and innovators ready to shape the futurefrom their desktop. Shop now for your favorite innovators: #SparkSomethingBig ✨ https://marketplace.nvidia.com/en-us/enterprise/personal-ai-supercomputers/ncid=so-twit-873553-vt33 https://marketplace.nvidia.com/en-us/enterprise/personal-ai-supercomputers/ncid=so-twit-873553-vt33"  
[X Link](https://x.com/NVIDIAAIDev/status/1999267203556712700)  2025-12-11T23:57Z 87.5K followers, [----] engagements


"🖼✨ How to achieve immersive photorealistic scene exploration in real-time From capture to streaming to display: Play4D from #NVIDIAResearch brings 4D scenes into Isaac Sim for Physical-AI testing in bespoke reconstructed environments with true free-viewpoint exploration. Discover how we're expanding access to 3D video ➡ 📹 by Prism AI https://research.nvidia.com/labs/amri/projects/play4d/ https://research.nvidia.com/labs/amri/projects/play4d/"  
[X Link](https://x.com/NVIDIAAIDev/status/1999283814179197381)  2025-12-12T01:03Z 86.8K followers, [----] engagements


"Find us at #SIGGRAPHAsia2025 to learn more about Play4D from NVIDIA Research. ➡ 📅 December 15-18 in 📍 Hong Kong https://sa2025.conference-schedule.org/presentation/id=emt_131&sess=sess183 https://sa2025.conference-schedule.org/presentation/id=emt_131&sess=sess183"  
[X Link](https://x.com/NVIDIAAIDev/status/1999283819178836371)  2025-12-12T01:03Z 86.8K followers, [----] engagements


"DGX Spark Live: Installation & First Run Guide for Robotics Developers https://x.com/i/broadcasts/1BdxYZOkWdyKX https://x.com/i/broadcasts/1BdxYZOkWdyKX"  
[X Link](https://x.com/NVIDIAAIDev/status/1999554952906371448)  2025-12-12T19:00Z 87.5K followers, [----] engagements


"Our team had a great time at #NeurIPS2025 last week including Ming-Yu Liu VP of Research at NVIDIA who had a session about NVIDIA Cosmos and discussed how the vision for world foundation models should be the matrix for robots 🤖 Watch his take on whats new in [----] for physical AI 👀 Also be sure to try out our new cookbook a practical guide to the Cosmos open models to help you get started with Cosmos ➡ https://nvidia-cosmos.github.io/cosmos-cookbook/ https://nvidia-cosmos.github.io/cosmos-cookbook/"  
[X Link](https://x.com/NVIDIAAIDev/status/1999615751838126202)  2025-12-12T23:02Z 87.3K followers, [----] engagements


"The Future You Can Gift -- For the innovator who has everything except this: 🎁 This holiday season give more than a gadget -- gift the gateway to innovation. NVIDIA DGX Spark empowers creators students and researchers to explore AI hands-on using our industry-standard frameworks. From experiment to execution its the AI gift that sparks possibility. 🛒 Shop now: #SparkSomethingBig ✨ https://nvda.ws/44XiLvf https://nvda.ws/44XiLvf"  
[X Link](https://x.com/NVIDIAAIDev/status/1999887038028324866)  2025-12-13T17:00Z 87.5K followers, 10.1K engagements


"Nemotron [--] Nano is now leading its size class on the latest Artificial Analysis leaderboards combining strong intelligence high openness and blazing output speed in a compact package. Part of new NVIDIA Nemotron [--] family Nemotron [--] Nano is powered by various advanced techniques including hybrid MoE architecture 1M context window and multi-environment RL training to achieve the leading accuracy and efficiency. Read the details in our tech blog: https://developer.nvidia.com/blog/inside-nvidia-nemotron-3-techniques-tools-and-data-that-make-it-efficient-and-accurate NVIDIA has just released"  
[X Link](https://x.com/NVIDIAAIDev/status/2000603340137435257)  2025-12-15T16:26Z 87.2K followers, 23.3K engagements


"Model Financial Risk like a Quant - The case for CVaR. https://developer.nvidia.com/blog/accelerating-real-time-financial-decisions-with-quantitative-portfolio-optimization/ https://developer.nvidia.com/blog/accelerating-real-time-financial-decisions-with-quantitative-portfolio-optimization/"  
[X Link](https://x.com/NVIDIAAIDev/status/2000974201746817278)  2025-12-16T17:00Z 87.7K followers, [----] engagements


"Scale up and supercharge financial portfolio optimization workflows using our NVIDIA CUDA-X libraries to reduce re-optimization from hours to seconds. ✅ Accelerate scenario generation up to 100x using cuML for KDE on 100k scenarios. ✅ Boost Mean-CVaR optimization cycles with cuOpt open-source solvers for up to 160x faster than state-of-the-art open-source CPU solvers at scale. Get started -- enroll in NVIDIAs new DLI Accelerating Portfolio Optimization hands-on course: 📝 Master the full optimization workflow: data curation to backtesting. 📝 Delve into accelerated optimization on NVIDIA GPUs"  
[X Link](https://x.com/NVIDIAAIDev/status/2001065049683574808)  2025-12-16T23:01Z 87.6K followers, [----] engagements


"The rise of AI factories physical AI and agents reshaped how AI was built and deployed in [----]. These are the advances that mattered most to developers ➡ https://developer.nvidia.com/blog/ai-factories-physical-ai-and-advances-in-models-agents-and-infrastructure-that-shaped-2025/ https://developer.nvidia.com/blog/ai-factories-physical-ai-and-advances-in-models-agents-and-infrastructure-that-shaped-2025/"  
[X Link](https://x.com/NVIDIAAIDev/status/2001080399376552024)  2025-12-17T00:02Z 87K followers, [----] engagements


"AI Learning Wrapped for the Holidays 🎁 Unwrap the power of AI this holiday season. NVIDIA DGX Spark bridges education and innovation helping students and developers build real-world skills in AI development and research -- all with NVIDIAs trusted software stack. 🛒 Shop now for your favorite innovator: #SparkSomethingBig ✨ https://marketplace.nvidia.com/en-us/enterprise/personal-ai-supercomputers/ncid=so-twit-873553-vt33 https://marketplace.nvidia.com/en-us/enterprise/personal-ai-supercomputers/ncid=so-twit-873553-vt33"  
[X Link](https://x.com/NVIDIAAIDev/status/2001324688518094923)  2025-12-17T16:12Z 87.5K followers, [----] engagements


"Starting the day seeing NVIDIA Nemotron [--] Nano trending at #1 on @huggingface🤗 Huge thank you to our developer community Tell us what you are building with it. https://nvda.ws/3Mzz3V1 https://nvda.ws/3Mzz3V1"  
[X Link](https://x.com/NVIDIAAIDev/status/2001351939892875504)  2025-12-17T18:01Z 87.5K followers, [----] engagements


"Join our latest hackathon winners - Team Tabasco including @rostammahabadi - for a live demo of OnSight AI a real-time AI safety and compliance system built during the NVIDIA DGX Spark Hackathon. Watch our next DGX Spark Livestream here or on YouTube: 🎥 📆 Friday December [--] 🕚 [--] a.m. Pacific #SparkSomethingBig https://youtube.com/live/t-GIii3WE9I https://youtube.com/live/t-GIii3WE9I https://youtube.com/live/t-GIii3WE9I https://youtube.com/live/t-GIii3WE9I"  
[X Link](https://x.com/NVIDIAAIDev/status/2001381890423443517)  2025-12-17T20:00Z 87.5K followers, [----] engagements


"Generative AI models can significantly improve semiconductor defect classification. For wafer map inspection: Fine-tuning Cosmos Reason VLM on wafer map defect data increases accuracy from zero-shot levels to over 96%. For die-level inspection: NV-DINOv2 fine-tuned with self-supervised learning (SSL) using NVIDIA TAO on unlabeled images achieves up to 8.9% accuracy improvement versus without SSL. Read the blog to get started with step-by-step instructions and a sample dataset ➡ https://nvda.ws/44xy5P4 https://nvda.ws/44xy5P4"  
[X Link](https://x.com/NVIDIAAIDev/status/2001457384938516514)  2025-12-18T01:00Z 87.3K followers, [----] engagements


"Predict LLM Performance with Dynamo AI Configurator https://x.com/i/broadcasts/1rmGPvepYrYxN https://x.com/i/broadcasts/1rmGPvepYrYxN"  
[X Link](https://x.com/NVIDIAAIDev/status/2001826948658671676)  2025-12-19T01:28Z 87.3K followers, [----] engagements


"The NVIDIA Nemotron family just crossed 5M downloads on @huggingface 🤗 A massive thank you to the community for your work and enthusiasm. 🏗 Get started here: https://nvda.ws/4pNTvQE https://nvda.ws/4pNTvQE"  
[X Link](https://x.com/NVIDIAAIDev/status/2002046308127842434)  2025-12-19T16:00Z 87.6K followers, 21K engagements


"DGX Spark Live: Real-Time Worker Safety with OnSight AI Powered by DGX Spark and Smart Glasses https://x.com/i/broadcasts/1yNxabvdZNjKj https://x.com/i/broadcasts/1yNxabvdZNjKj"  
[X Link](https://x.com/NVIDIAAIDev/status/2002091435953463658)  2025-12-19T18:59Z 87.5K followers, [----] engagements


"📣 Were collaborating with @awscloud to help developers move AI agents from prototype to production. ✅ Build and customize foundation models with Strands Agents and NVIDIA NeMo ✅ Optimize agent workflows with the NVIDIA NeMo Agent Toolkit ✅ Deploy and operate agents securely using Amazon Bedrock AgentCore Together were enabling scalable production-ready agentic AI 👉 https://nvda.ws/49h5VdT https://nvda.ws/49h5VdT"  
[X Link](https://x.com/NVIDIAAIDev/status/2002136864006823971)  2025-12-19T22:00Z 87.5K followers, 12.5K engagements


"Reservoir simulation is vital for forecasting and field planning but thousands of runs remain costly even with GPU acceleration. XMeshGraphNet offers a fast surrogate model aligned with finite-volume methods handling complex geology and scaling to million-cell industry-scale reservoirs. ➡ https://github.com/NVIDIA/physicsnemo/tree/main/examples/reservoir_simulation/xmgn https://github.com/NVIDIA/physicsnemo/tree/main/examples/reservoir_simulation/xmgn"  
[X Link](https://x.com/NVIDIAAIDev/status/2002786140529508715)  2025-12-21T17:00Z 87.5K followers, [----] engagements


"Last chance to save 50% on NVIDIA certification exams. Offer ends December [--]. Earn credentials that set you apart in generative and agentic AI OpenUSD accelerated data science and AI infrastructure and operations. Validate your skills and start the new year NVIDIA-Certified. 🎉 https://nvda.ws/44IG2kx https://nvda.ws/44IG2kx"  
[X Link](https://x.com/NVIDIAAIDev/status/2003188764030415112)  2025-12-22T19:39Z 87.6K followers, [----] engagements


"👀 New Skip Softmax a hardware-friendly drop-in sparse attention method that accelerates #inference without any retraining. Learn how Skip Softmax delivers up to 1.4x faster time-to-first-token (TTFT) and up to 1.4x faster time-per-output-token (TPOT) and how to get started with FlashAttention and NVIDIA TensorRT LLM. 🏎 Tech blog ➡ https://developer.nvidia.com/blog/accelerating-long-context-inference-with-skip-softmax-in-nvidia-tensorrt-llm/ https://developer.nvidia.com/blog/accelerating-long-context-inference-with-skip-softmax-in-nvidia-tensorrt-llm/"  
[X Link](https://x.com/NVIDIAAIDev/status/2003526015981654160)  2025-12-23T18:00Z 87.7K followers, [----] engagements


"Huge shoutout to @UnslothAI for being an incredible partner powering fast efficient fine-tuning of Nemotron [--] and enabling reinforcement learning with NVIDIA NeMo Gym integration. NVIDIA Nemotron [--] Nano - How To Run Guide ➡ https://docs.unsloth.ai/models/nemotron-3 NVIDIA made a beginner's guide to fine-tuning LLMs with Unsloth 💚 You'll learn about: - Training methods: LoRA FFT RL - When to fine-tune and why + use-cases - Amount of data and VRAM needed - How to train locally on DGX Spark RTX GPUs & more Guide: https://t.co/ajwL99Bug7 https://t.co/D0vyJkza0B"  
[X Link](https://x.com/NVIDIAAIDev/status/2003541323966795792)  2025-12-23T19:00Z 87.6K followers, 17.3K engagements


"Top [--] AI Model Optimization Techniques for Faster Smarter Inference [--] Post-Training Quantization (PTQ) The fastest path to value. Compress models without retraining for instant latency and throughput wins. [--] Quantization-Aware Training (QAT) Fine-tune for low precision and recover accuracy when PTQ isnt enough. [--] Quantization-Aware Distillation (QAD) Go further by pairing quantization with distillation to unlock ultra-low-precision inference with maximum quality. [--] Speculative Decoding Slash decode latency by collapsing sequential stepsno retraining required. [--] Pruning + Distillation"  
[X Link](https://x.com/NVIDIAAIDev/status/2003572784677396653)  2025-12-23T21:05Z 87.7K followers, 34.8K engagements


"Happy Holidays Thanks for building the future with us. 💚 NVIDIA dev team"  
[X Link](https://x.com/NVIDIAAIDev/status/2003828006058373579)  2025-12-24T14:00Z 87.7K followers, 10.7K engagements


"Build and deploy your own autonomous deep research agent using NVIDIA's AI-Q Research Assistant Blueprint in this new self-paced course. You'll learn how to: ✅ Deploy NVIDIA Nemotron reasoning models ✅ Build a RAG pipeline with vector databases ✅ Orchestrate agentic workflows with human-in-the-loop capabilities Enroll now: https://nvda.ws/4sjcyUD https://nvda.ws/4sjcyUD"  
[X Link](https://x.com/NVIDIAAIDev/status/2006401935570509908)  2025-12-31T16:27Z 88.5K followers, 31K engagements


"We're counting down the hours until NVIDIA Live at #CES2026 📣 Join us at [--] p.m. PT as our Founder and CEO Jensen Huang shares whats next in AI in a special presentation. 🔗 https://nvda.ws/3YscCUj https://nvda.ws/3YscCUj"  
[X Link](https://x.com/NVIDIAAIDev/status/2008233086367011019)  2026-01-05T17:44Z 89.1K followers, 40.5K engagements


"Coming up next on the NVIDIA Live pregame 🔗 Tune in: https://www.youtube.com/watchv=0NBILspM4c4 CodeRabbit CEO Harjot Gill will be joining the NVIDIA Live pregame as a panelist to discuss how open models help startups move faster and scale globally Tune in today at 11:45 a.m. PT before NVIDIA CEO Jensen Huang takes the stage https://t.co/ZHGio7phf6 https://www.youtube.com/watchv=0NBILspM4c4 CodeRabbit CEO Harjot Gill will be joining the NVIDIA Live pregame as a panelist to discuss how open models help startups move faster and scale globally Tune in today at 11:45 a.m. PT before NVIDIA CEO"  
[X Link](https://x.com/NVIDIAAIDev/status/2008268511295271344)  2026-01-05T20:04Z 88.4K followers, [----] engagements


"NVIDIA Cosmos Reason [--] is here. 🥳 An open highly accurate reasoning vision language model for physical AI featuring: ✅ Improved spatio-temporal understanding and timestamp precision ✅ Flexible deployment with 2B and 8B model sizes ✅ Long-context reasoning with up to 256K tokens ✅ Expanded visual perception across complex environments We also have new Cosmos releases: Predict [---] Transfer [---] and the NVIDIA GR00T N1.6 robot foundation model. 📗Read our technical blog: 🤗 Download Cosmos Reason [--] on @HuggingFace: https://nvda.ws/3L4B6Qy https://nvda.ws/4swwC68 https://nvda.ws/3L4B6Qy"  
[X Link](https://x.com/NVIDIAAIDev/status/2008313146524520767)  2026-01-05T23:02Z 89.4K followers, 45.8K engagements


"Turn @HuggingFace Reachy Mini into your own AI assistant. 🤖 New step-by-step guide walks you through creating an interactive agent using Nemotron [--] open models Brev and DGX Spark. Bring Reachy Mini to life running locally and in the cloud. 🤗 Read the tutorial: https://huggingface.co/blog/nvidia-reachy-mini Super cool to see Jensen @nvidia showcasing Reachy Mini at his #CES26 keynote. Paired with a DGX Spark & Brev it can make the perfect local home AI robotics setup https://t.co/DkvbqrcXoU https://huggingface.co/blog/nvidia-reachy-mini Super cool to see Jensen @nvidia showcasing Reachy Mini"  
[X Link](https://x.com/NVIDIAAIDev/status/2008319790868201795)  2026-01-05T23:28Z 89.4K followers, 23.6K engagements


"Run 1T parameter models at your desk 🤯 At #CES2026 we unveiled how developers can run the latest open and frontier AI models on a local system from 100B parameter models on DGX Spark to 1T parameter models on DGX Station. Read the blog ➡ https://nvda.ws/4q8dee1 https://nvda.ws/4q8dee1"  
[X Link](https://x.com/NVIDIAAIDev/status/2008327654265565335)  2026-01-06T00:00Z 89.3K followers, 26.3K engagements


"NVIDIA DGX Spark keeps getting faster. ✨⚡ New software models and open-source optimizations deliver up to [---] higher performance with NVFP4 run 235B-parameter models locally and allow multiple workloads to run simultaneously on the desktop. Read the technical blog ➡ https://nvda.ws/45igTh5 https://nvda.ws/45igTh5"  
[X Link](https://x.com/NVIDIAAIDev/status/2008335204167901562)  2026-01-06T00:30Z 89.9K followers, 10.1K engagements


"🤯 100% of NVIDIA engineers code with AIand theyre checking in 3x more code than before. At that scale human-only code review cant keep up. @coderabbitai review agents use models like Claude and GPT with NVIDIA Nemotron to: ✅ Pull context from code docs project trackers and more ✅ Use Nemotrons long context and reasoning to summarize ✅ so that frontier models can flag issues and suggest fixes in minutes not days. Join us today at [--] AM PT to see an end-to-end demo and bring your Qs: 🤗 Get Nemotron Nano 3: https://nvda.ws/44TT138 https://nvda.ws/49mVcgV https://nvda.ws/44TT138"  
[X Link](https://x.com/NVIDIAAIDev/status/2008569752956854592)  2026-01-06T16:02Z 89.6K followers, 100.3K engagements


"@stephsmithio Welcome to team green 💚"  
[X Link](https://x.com/NVIDIAAIDev/status/2008584008377262197)  2026-01-06T16:58Z 88.5K followers, [---] engagements


"Extreme co-design is the foundation of the NVIDIA Rubin platform. For anyone who really wants to go deep we put together an under-the-hood deep dive. Fair warning: its a 60-minute read. Learn how compute networking power delivery cooling and more are architected together as a single system for scalable intelligence production 👉 https://nvda.ws/4pvGytQ https://nvda.ws/4pvGytQ"  
[X Link](https://x.com/NVIDIAAIDev/status/2008600710510764211)  2026-01-06T18:05Z 88.6K followers, [----] engagements


"Physical AI needs models that truly understand the real world. The NVIDIA Cosmos platform addresses this challenge with open world foundation models purpose-built for human-like reasoning and world generation. From simulation to autonomy Cosmos world models accelerate how robots and AI agents learn to reason plan and act. Announcing open high accuracy and customizable models like Cosmos Reason [--] Cosmos Predict [---] and Cosmos Transfer [---] that are pushing the boundaries of embodied intelligence. Learn more 📖 Explore the Cosmos Cookbook 🧑🍳 https://nvda.ws/3YWgEV8 https://nvda.ws/4stzmB5"  
[X Link](https://x.com/NVIDIAAIDev/status/2008629645441184227)  2026-01-06T20:00Z 88.8K followers, [----] engagements


"Just launched #CES2026 the new open-source NVIDIA Nemotron Speech ASR model is here to solve latency drift and redundant compute. Its cache-aware streaming architecture eliminates the need for buffered inference giving you stable sub-100ms latency (24ms median T-T-F) and up to 3x more throughput on your GPU. 🤗 Read the technical blog with real-world results from @trydaily and @modal on @HuggingFace: https://nvda.ws/3Lt8m3Q https://nvda.ws/3Lt8m3Q"  
[X Link](https://x.com/NVIDIAAIDev/status/2008654492204441862)  2026-01-06T21:38Z 90.1K followers, 137.7K engagements


"@Fortinet @nvidia 💚"  
[X Link](https://x.com/NVIDIAAIDev/status/2008681919698137439)  2026-01-06T23:27Z 89.1K followers, [---] engagements


"@CrowdStrike @nvidia Excited to see the progress in applying AI agents to security operations We look forward to how our continued partnership helps teams detect investigate and protect faster"  
[X Link](https://x.com/NVIDIAAIDev/status/2008685563810041896)  2026-01-06T23:42Z 89.3K followers, [---] engagements


"At #CES2026 we unveiled [--] new open modelsspanning autonomous driving robotics healthcare and agentic AI. In this video @Sam_Witteveen breaks it all down: ✅ The NVIDIA Rubin Platform ✅ Alpamayo for self-driving cars ✅ Nemotron Speech ASR for voice agents ✅ Cosmos open world foundation models to advance physical AI Watch here👇 https://nvda.ws/3LoSHmi https://nvda.ws/3LoSHmi"  
[X Link](https://x.com/NVIDIAAIDev/status/2008961836121526641)  2026-01-07T18:00Z 89.4K followers, [----] engagements


"Build a lightning-fast voice agent with NVIDIA Nemotron Speech ASRnow powering @trydaily's latest low-latency voice experience with voice-to-voice latency under [---] ms. Check out how Daily uses NVIDIA open models to deliver responsive scalable voice agents for real-time conversations in their new blog 👇 Build a voice agent with the brand new @nvidia Nemotron Speech ASR open source transcription model and @pipecat_ai. This model was designed from the ground up for low-latency use cases like voice agents. It performs very well on our benchmarks and is a new state of the art for"  
[X Link](https://x.com/NVIDIAAIDev/status/2009020320259223902)  2026-01-07T21:52Z 89.4K followers, 10.5K engagements


"Live from #CES 🎉 Dynamo Office Hours are here. Learn how NVIDIA Dynamo integrates with Kubernetes via the Gateway API and Inference Extension and get direct support from our experts. Perfect for beginners and power users alike. Tune into office hours: Continue the journey at Dynamo Day: https://nvda.ws/45wbiDI https://nvda.ws/49pkL0O https://nvda.ws/45wbiDI https://nvda.ws/49pkL0O"  
[X Link](https://x.com/NVIDIAAIDev/status/2009041104746721414)  2026-01-07T23:15Z 89.3K followers, [----] engagements


"New Cosmos Cookbook recipe just dropped 🍽 @intbotais recipe evaluates Cosmos Reason [--] on a set of targeted egocentric video tests designed to probe robot-relevant reasoning including: 📖 Human intent directed toward the robot 📖 Object motion relative to the robot 📖 Spatial relationships among multiple people 📖 Social context awareness (when not to engage) Try it yourself in the Cosmos Cookbook 👉 Download Cosmos Reason [--] 👉 https://nvda.ws/45JQWa6 https://nvda.ws/3N6bfbz https://nvda.ws/45JQWa6 https://nvda.ws/3N6bfbz"  
[X Link](https://x.com/NVIDIAAIDev/status/2009309122521423882)  2026-01-08T17:00Z 89.3K followers, [----] engagements


"Live from #CES2026 Spark hits the show floor ✨ Thanks to everyone for stopping by to meet Reachy Mini an AI agent brought to life right at your desk powered by DGX Spark. #SparkSomethingBig"  
[X Link](https://x.com/NVIDIAAIDev/status/2009414820425015674)  2026-01-09T00:00Z 89.5K followers, [----] engagements


"CES meets DGX Spark [----] wrap-up https://x.com/i/broadcasts/1yNxabBYezrKj https://x.com/i/broadcasts/1yNxabBYezrKj"  
[X Link](https://x.com/NVIDIAAIDev/status/2009701447324279063)  2026-01-09T18:58Z 89.5K followers, [----] engagements


"Join the Dynamo community and ecosystem to explore the latest features in LLM inference KV-cache optimization and distributed serving. Gain insights from leading OSS inference framework maintainers renowned inference thought leaders real-world inference deployments and Dynamo users building high-performance inference pipelines. 📅 Thursday Jan [--] [--] AM1 PM PT 🎟 Sign up for Dynamo Day: https://nvda.ws/4jxCxnl https://nvda.ws/4jxCxnl"  
[X Link](https://x.com/NVIDIAAIDev/status/2009716818147971547)  2026-01-09T20:00Z 89.2K followers, [----] engagements


"New Nemotron RAG multimodal embed + re-rank models. Production-ready VLMs for high-accuracy document retrieval across PDFs charts tables and slides. 🤗 Check out our blog on @HuggingFace: https://nvda.ws/3NmdabU https://nvda.ws/3NmdabU"  
[X Link](https://x.com/NVIDIAAIDev/status/2009762337738183100)  2026-01-09T23:00Z 89.5K followers, 14.7K engagements


"@WOLF_Financial 🙌"  
[X Link](https://x.com/NVIDIAAIDev/status/2010458752802308429)  2026-01-11T21:08Z 89.5K followers, [---] engagements


"Start the new year right with NVIDIA Deep Learning Institute for virtual workshops designed to develop key skills in AI accelerated data science or accelerated computing. Check out the new schedule with workshops in model parallelism multimodal models and RAG pipelines. Don't just watch a video. Build it live with a DLI-certified instructor. Check the schedule and register now: https://nvda.ws/45JUXLF https://nvda.ws/45JUXLF"  
[X Link](https://x.com/NVIDIAAIDev/status/2010788880161608191)  2026-01-12T19:00Z 89.5K followers, [----] engagements


"Transform warehouse operations with multiagent intelligence 🤖📦 Most warehouses are drowning in data but starving for decisions. The new Multi-Agent Intelligent Warehouse command layer built on NVIDIAs stack mirrors real warehouse roles eliminating fragmentation and tribal knowledge without ripandreplace or vendor lockin. Read the tech blog: Explore the full blueprint: https://nvda.ws/4pFL7Sq https://nvda.ws/45DQ9HM https://nvda.ws/4pFL7Sq https://nvda.ws/45DQ9HM"  
[X Link](https://x.com/NVIDIAAIDev/status/2010849273563210228)  2026-01-12T23:00Z 89.4K followers, [----] engagements


"@rdesh26 @nvidia Excited to have you join the team - welcome"  
[X Link](https://x.com/NVIDIAAIDev/status/2010954002008326655)  2026-01-13T05:56Z 89.7K followers, [---] engagements


"Its now easier than ever for developers to start building with instant access to GPU instances through NVIDIA Brev on popular cloud platforms like @shadeformai. Launch customize and deploy AI models and applications in just a few clicks. Today were super excited to announce @NVIDIA has integrated @shadeformai 's compute marketplace to supply NVIDIA GPUs to the NVIDIA Brev platform https://t.co/a0X4WQSWgc Today were super excited to announce @NVIDIA has integrated @shadeformai 's compute marketplace to supply NVIDIA GPUs to the NVIDIA Brev platform https://t.co/a0X4WQSWgc"  
[X Link](https://x.com/NVIDIAAIDev/status/2011136398104150100)  2026-01-13T18:00Z 89.1K followers, [----] engagements


"Agentic AI Solutions with CrewAI and Nemotron https://x.com/i/broadcasts/1lPJqvXZEDNxb https://x.com/i/broadcasts/1lPJqvXZEDNxb"  
[X Link](https://x.com/NVIDIAAIDev/status/2011151536605577382)  2026-01-13T19:01Z 89.7K followers, [----] engagements


"Its now easier than ever for developers to start building with instant access to NVIDIA GPU instances on popular cloud platforms through NVIDIA Brev with @shadeformai. Launch customize and deploy AI models and applications in just a few clicks. Today were super excited to announce @NVIDIA has integrated @shadeformai 's compute marketplace to supply NVIDIA GPUs to the NVIDIA Brev platform https://t.co/a0X4WQSWgc Today were super excited to announce @NVIDIA has integrated @shadeformai 's compute marketplace to supply NVIDIA GPUs to the NVIDIA Brev platform https://t.co/a0X4WQSWgc"  
[X Link](https://x.com/NVIDIAAIDev/status/2011161378162426333)  2026-01-13T19:40Z 89.7K followers, [----] engagements


""We no longer have to train an AI model to know everything on day one." AI isnt just predicting the next word anymore. Its reasoning through problems its never seen.💡 Watch the full keynote: https://nvda.ws/4qsDhNi https://nvda.ws/4qsDhNi"  
[X Link](https://x.com/NVIDIAAIDev/status/2011196560751681814)  2026-01-13T22:00Z 89.8K followers, 15.6K engagements


"@Ahmad_Al_Dahle Congrats on the next chapter"  
[X Link](https://x.com/NVIDIAAIDev/status/2011456244071219470)  2026-01-14T15:11Z 89.5K followers, [----] engagements


"@joaomdmoura she's headed places"  
[X Link](https://x.com/NVIDIAAIDev/status/2011496685344866504)  2026-01-14T17:52Z 89.5K followers, [---] engagements


"GTC [----] registration is now live. Choose from nine full-day instructor-led workshops covering topics in deep learning AI infrastructure LLMs physical AI and more. Workshops at past GTCs were completely sold out and demand is high for [----]. This is your opportunity to get hands-on training onsite. Validate your skills with certification at GTC too. Register now: https://nvda.ws/45cFDqP https://nvda.ws/45cFDqP"  
[X Link](https://x.com/NVIDIAAIDev/status/2011513648938697036)  2026-01-14T19:00Z 89.5K followers, [----] engagements


"Why store every KV pair when the model only attends to a fraction of them ⚡Meet KVzap: Smart input-dependent KV cache pruning. By learning to discard specific cache entries that won't be needed for future lookups it cuts cache size by 2-4 compression with negligible accuracy loss. It achieves state-of-the-art performance on the KVpress Leaderboard. 📝 Code: 📗 Paper: 🏆 Leaderboard: https://huggingface.co/spaces/nvidia/kvpress-leaderboard https://arxiv.org/abs/2601.07891 https://github.com/NVIDIA/kvpress/tree/main/kvzap We are open-sourcing KVzap a SOTA KV cache pruning method that delivers"  
[X Link](https://x.com/NVIDIAAIDev/status/2011528243413651539)  2026-01-14T19:58Z 89.5K followers, 37.5K engagements


"@googledevs @nvidia @googlecloud 🙌"  
[X Link](https://x.com/NVIDIAAIDev/status/2011631884695204125)  2026-01-15T02:49Z 89.5K followers, [---] engagements


"Ready to take your Bash agent from handy to specialized Our newest blog will help you teach your Bash Agent a brand new CLI tool as well as outline how you could add your own tool. [--] Use NeMo Data Designer to.design your synthetic data. [--] Send your Agent to the Reinforcement Learning Gym with @UnslothAI [--] Run the agent with the new skill Read ➡ https://nvda.ws/3LOnLMm https://nvda.ws/3LOnLMm"  
[X Link](https://x.com/NVIDIAAIDev/status/2011847842604990647)  2026-01-15T17:07Z 89.5K followers, [----] engagements


"Disaggregated serving breaks one GPU does everything monoliths into purpose-built specialization. By splitting model inference into different stages and running them on separate GPU pools every stage gets optimized independently and gets the exact resources it needs. Watch the full explanation 👇 Learn more: https://nvda.ws/3LCIrqG https://nvda.ws/3LCIrqG"  
[X Link](https://x.com/NVIDIAAIDev/status/2011923894039372258)  2026-01-15T22:10Z 89.5K followers, [----] engagements


"Get firsthand insights on disaggregated serving from the Dynamo community at the upcoming Dynamo Day. https://nvda.ws/45aOk51 https://nvda.ws/45aOk51"  
[X Link](https://x.com/NVIDIAAIDev/status/2011923897864601956)  2026-01-15T22:10Z 89.6K followers, [----] engagements


"History made 🏆 NVIDIA cuOpt just won the [----] LPFeas Hans Mittelmann leaderboard becoming the worlds fastest open-source LP solver. The GPU-accelerated chain reaction in decision optimization is here with three out of the four LP solvers now GPU-accelerated. Get started: https://nvda.ws/49J984Z https://nvda.ws/49J984Z"  
[X Link](https://x.com/NVIDIAAIDev/status/2011951538449559695)  2026-01-16T00:00Z 89.5K followers, [----] engagements


"Generalist models give agents range. Specialized models give them expertise. NVIDIA's VP of Generative AI Software @karibriski sat down with @amitisinvesting to discuss how open models like NVIDIA Nemotron help you win the "last mile of AI." Watch the full interview 👇 https://nvda.ws/4qmW7Ff https://nvda.ws/4qmW7Ff"  
[X Link](https://x.com/NVIDIAAIDev/status/2012306948624888256)  2026-01-16T23:32Z 90K followers, 24.6K engagements


"We noticed one of the GPU-accelerated solvers was missing from the comparison. Weve updated our graph to include HPRLP"  
[X Link](https://x.com/NVIDIAAIDev/status/2012309324576506100)  2026-01-16T23:41Z 89.7K followers, [----] engagements


"@mrdbourke @huggingface 🙌 🤗"  
[X Link](https://x.com/NVIDIAAIDev/status/2013450313404125237)  2026-01-20T03:15Z 89.5K followers, [---] engagements


"Learn with Me: Train AI Agents for Command-Line Tasks with Synthetic Data and RL Nemotron Labs https://x.com/i/broadcasts/1dRJZaleqmrGB https://x.com/i/broadcasts/1dRJZaleqmrGB"  
[X Link](https://x.com/NVIDIAAIDev/status/2013687872570990719)  2026-01-20T18:59Z 89.7K followers, [----] engagements


"Congrats to @StepFun_ai on the release of Step-Audio-R1.1 and the #1 leaderboard result. Step-Audio-R1.1 is optimized for NVIDIA GPUs and designed for real-time high-accuracy voice agents. Check it out 👉 https://huggingface.co/stepfun-ai/Step-Audio-R1.1 StepFuns new Step-Audio R1.1 (Realtime) is the new leading Speech to Speech model surpassing Grok Voice Agent in our Big Bench Audio benchmark The new model from StepFun achieves a score of 96.4% on Big Bench Audio ahead of the previous leader xAIs Grok Voice Agent. StepFun https://t.co/xdi9Y64aOr"  
[X Link](https://x.com/NVIDIAAIDev/status/2014039326989144551)  2026-01-21T18:16Z 89.6K followers, [----] engagements


"@PyTorch Thanks for sharing 🙌"  
[X Link](https://x.com/NVIDIAAIDev/status/2014099383248171144)  2026-01-21T22:14Z 89.7K followers, [---] engagements


"@ArmSoftwareDev @nvidia @Dell Thanks for joining"  
[X Link](https://x.com/NVIDIAAIDev/status/2014336287713669454)  2026-01-22T13:56Z 89.7K followers, [---] engagements


"What city should we stop by next for a dev meetup 🚐 Our crew from @NVIDIAAIDev with @Baxate crushing our first of the new monthly NVIDIA Bay area developer meetups. Celebrating the work of innovators builders and doers. We will be coming to a major city near you in [----] as well 💚🤙😎 https://t.co/7jiKeF4AxL Our crew from @NVIDIAAIDev with @Baxate crushing our first of the new monthly NVIDIA Bay area developer meetups. Celebrating the work of innovators builders and doers. We will be coming to a major city near you in [----] as well 💚🤙😎 https://t.co/7jiKeF4AxL"  
[X Link](https://x.com/NVIDIAAIDev/status/2014715305457811735)  2026-01-23T15:02Z 89.9K followers, [----] engagements


"Just released by @NVIDIA on GitHub: nv-wavenet a reference implementation of a #CUDA-enabled autoregressive WaveNet inference engine. https://nvda.ws/2vBt6il https://nvda.ws/2vBt6il"  
[X Link](https://x.com/NVIDIAAIDev/status/988514217321009152)  2018-04-23T20:25Z 89.2K followers, [---] engagements

Limited data mode. Full metrics available with subscription: lunarcrush.com/pricing