[GUEST ACCESS MODE: Data is scrambled or limited to provide examples. Make requests using your API key to unlock full data. Check https://lunarcrush.ai/auth for authentication information.]

[@MVanbrunsc47513](/creator/twitter/MVanbrunsc47513)
"I think its going to be a hybrid model a developerowner kind of partnership. Poolside keeps control of land power and capacity while modular EPCs and CoreWeave handle execution. This makes the most sense to me. The 3-person team is enough for oversight because prefab modules outsourced EPC and leaseback operators make it possible to deliver a multi-hundred-MW campus with minimal internal staff. They can scale this a bit later on when needed It looks like the start of a broader trend of model labs moving below the IaaS layer to own land power and interconnect rights while leasing early"  
[X Link](https://x.com/MVanbrunsc47513/status/1978615232902693334) [@MVanbrunsc47513](/creator/x/MVanbrunsc47513) 2025-10-16T00:13Z 2097 followers, 3977 engagements


"@JJPeraltaa Thanks man🙂"  
[X Link](https://x.com/MVanbrunsc47513/status/1972734722410508415) [@MVanbrunsc47513](/creator/x/MVanbrunsc47513) 2025-09-29T18:46Z 2089 followers, XX engagements


"A follow-up to my post last week covering Jeff Bezoss interview at Italian Tech Week were already seeing more evidence supporting the statements he made. Bezos called the current AI cycle an industrial bubble a phase where excitement floods every company pitch. When people get very excited every experiment gets funded the good ideas and the bad ideas. Thats not an exaggeration. Over XX% of all VC money now goes into AI startups compared to XX% for internet companies at the height of the 1999 boom. As stated by Bezos were seeing six-person teams raising nine-digit rounds. Capital is chasing"  
[X Link](https://x.com/MVanbrunsc47513/status/1975155278657040874) [@MVanbrunsc47513](/creator/x/MVanbrunsc47513) 2025-10-06T11:04Z 2088 followers, 3824 engagements


"Back in March DataONE already mentioned BTM (behind the meter) power for the Vineland build. My guess based on what Im seeing across the landscape: on-site natural-gas simple-cycle turbines paired with possible utility-scale batteries. Makes perfect sense for the 24/7 AI load before other solutions kick in"  
[X Link](https://x.com/MVanbrunsc47513/status/1978188494976712921) [@MVanbrunsc47513](/creator/x/MVanbrunsc47513) 2025-10-14T19:57Z 2087 followers, 2062 engagements


"Keep in mind that you need a decent exit strategy on all these names in XX to X years. Hard to say depends on the curve. The real risk is how theyll retain the talent and knowledge. Their true moat isnt hardware or powerits scarcity and know-how. Thats where investor focus should stay on talent and knowledge retention"  
[X Link](https://x.com/MVanbrunsc47513/status/1978207618368667900) [@MVanbrunsc47513](/creator/x/MVanbrunsc47513) 2025-10-14T21:13Z 2088 followers, XXX engagements


"@FransBakker9812 @eisokant @CoreWeave It will be interesting to see because it will reveal a lot on how strong the knowledge edge is"  
[X Link](https://x.com/MVanbrunsc47513/status/1978507485133570476) [@MVanbrunsc47513](/creator/x/MVanbrunsc47513) 2025-10-15T17:05Z 2088 followers, XXX engagements


"@FransBakker9812 @eisokant @CoreWeave Eventually this will be the base imo but now already would be really soon"  
[X Link](https://x.com/MVanbrunsc47513/status/1978507718764777870) [@MVanbrunsc47513](/creator/x/MVanbrunsc47513) 2025-10-15T17:06Z 2088 followers, XXX engagements


"A few days back I already wrote a X part post on how potentially brittle the moat of companies like $NBIS $IREN $CRWV can be. It is all based on knowledge and power scarcity (temporary). This post by Poolside AI underwrited that risk. Links below on parameters too watch regarding these risks 👇"  
[X Link](https://x.com/MVanbrunsc47513/status/1978512074671030353) [@MVanbrunsc47513](/creator/x/MVanbrunsc47513) 2025-10-15T17:23Z 2097 followers, 10.7K engagements


"@TacticzH Like i said the fear and gread index should not be used that often. Its a lagging indicator and only slightly works in a certain bandwith. It mistakes emotion for volatilty. And decent investors dont use it. Maybe thats why it is all over X now 😅"  
[X Link](https://x.com/MVanbrunsc47513/status/1979347837700968732) [@MVanbrunsc47513](/creator/x/MVanbrunsc47513) 2025-10-18T00:44Z 2097 followers, XXX engagements


"@TacticzH Missing ASICS with broadcom/marvell. And I will also say missing some essential sectors who focus on optimalizing next gen data center with silicon test solutions and power/heat solving solutions"  
[X Link](https://x.com/MVanbrunsc47513/status/1975667708546195461) [@MVanbrunsc47513](/creator/x/MVanbrunsc47513) 2025-10-07T21:01Z 2095 followers, 1488 engagements


"@daniel_koss Nebius would do the right thing to get debt or equity dilution if the ROI on Clickhouse keeps growing and works to an exit event. It happens a lot so calling it the smartest move EVER is some headline bs. It is basic math and risk/reward skew"  
[X Link](https://x.com/MVanbrunsc47513/status/1976334554458910954) [@MVanbrunsc47513](/creator/x/MVanbrunsc47513) 2025-10-09T17:10Z 2095 followers, XXX engagements


"Good that you put the time into it Daniel. The structure makes sense and I like that you integrated risk. I do have substantial points Id tighten: The $24B total feels on the high end. ClickHouse was last valued around $6.35B (May XX Series C) with Nebius holding roughly XX% thats about $1.61.8B stake value today not $10B. I do feel there should be some middle ground since their excecution from May until now. The 40b IPO is just social media chatter and absolutely not base case. Theres upside if they IPO or re-rate but thats not realized yet. All other stakes are inflated to in that"  
[X Link](https://x.com/MVanbrunsc47513/status/1977849015585976334) [@MVanbrunsc47513](/creator/x/MVanbrunsc47513) 2025-10-13T21:28Z 2095 followers, 1166 engagements


"Thank you Oguz The thing is in my humble opinion its more wider and complex then you just framed. Also companies like $FLNC have been seeing a lot of retail capital flow in but Im not very impressed by their execution. Let alone margins. This phase is mostly focussed on chaos and opportunity chasing. Fluence had scaling issues before and the market is getting more packed every second. They became profitable but its a tough landscape. My earlier write-up on power volatility goes beyond grid volatility it looks at build-out volatility and long-term systemic load. A lot of companies are now"  
[X Link](https://x.com/MVanbrunsc47513/status/1978245273617936573) [@MVanbrunsc47513](/creator/x/MVanbrunsc47513) 2025-10-14T23:43Z 2095 followers, XXX engagements


"@FransBakker9812 @eisokant @CoreWeave Since now. They announced it on their blog didnt they"  
[X Link](https://x.com/MVanbrunsc47513/status/1978492095632413132) [@MVanbrunsc47513](/creator/x/MVanbrunsc47513) 2025-10-15T16:04Z 2088 followers, XXX engagements


"@FransBakker9812 @eisokant @CoreWeave Haha I thought you where serious for a second and mocking them"  
[X Link](https://x.com/MVanbrunsc47513/status/1978505562519208413) [@MVanbrunsc47513](/creator/x/MVanbrunsc47513) 2025-10-15T16:57Z 2088 followers, XXX engagements


"@TheKamaHsutra i think IRENs real moat starts with the grid permission layer. They already control XXX GW of reserved capacity and should actively pull that asset lever (which they are doing). thats basically a time monopoly. If they turn those queue positions into multi-tenant substations it becomes real monetizable infrastructure. From there step X should be imo push into efficiency. Something like high-density cooling with partners like Submer or Iceotope to build proprietary thermal IP and host multirome more GPUs per megawatt in efficiency. Then layer in smart power trading to cut"  
[X Link](https://x.com/MVanbrunsc47513/status/1979311860102107472) [@MVanbrunsc47513](/creator/x/MVanbrunsc47513) 2025-10-17T22:21Z 2097 followers, XXX engagements


"Well i agree from a tech hardware cadence perspective but it depends if Oracle sees it self as a tech leader or a legacy enterprise software company (and its investors). This move shows imo they strategic chose for the latter part. Oracle pays a regular quarterly dividend. That means cash generation and earnings stability matter more to its shareholders than hypergrowth. So management is incentivized to smooth earnings. And because their investor base is more focussed on that then people who invest in NVIDIA Coreweave etc they chose this strategy imo. The Wall Street optics matter more to"  
[X Link](https://x.com/MVanbrunsc47513/status/1979564048954868184) [@MVanbrunsc47513](/creator/x/MVanbrunsc47513) 2025-10-18T15:03Z 2097 followers, XXX engagements