What is true on Gpu?

We’ve spent the past year developing new primitives for GPUs. We believe this enables new algorithms to change how machines learn, open new markets, and dramatically reshape AI’s cost structure. We have raised a $1 billion Series A, co-led by Index and Lux. https://t.co/VUkQ68gL3U

Why hardware specialization is back

$SPCX just reported ABSOLUTELY INSANE earnings, and I don't think people realize how quickly Elon Musk is scaling this company. SpaceX generated a staggering $14.6 BILLION in quarterly revenue, representing 161% year over year growth, while earnings per share surged 197% to $0.35. That's a 12% revenue beat and a 119% EPS beat against the estimates shown. And the numbers get even CRAZIER. SpaceX reportedly generated $4.4B in net income and $8.8B in adjusted EBITDA, with EBITDA growing an incredible 306% year over year. We're talking about a company that's not only more than doubling its revenue but potentially expanding its profitability at an even faster pace.

This is honestly insane. Marc Boroditsky, CRO of $NBIS, just revealed that the company has 1 GW of Vera Rubin GPU capacity coming next year, with significant customer demand already lined up. That alone could translate into tens of billions of dollars in ARR. 🤯 https://t.co/X2CNpaJaSH

ELON MUSK TARGETS 13 GW OF AI COMPUTE BY 2027, WITH 440,000 NVIDIA VERA RUBIN GPUS IN A SINGLE TRAINING CLUSTER Musk says SpaceX is on track to bring roughly 200,000 $NVDA GB300 GPUs online per month for the remainder of 2026, with deployments continuing at a similar pace next year, primarily using Vera Rubin. The company expects to have 2.5-3 GW of compute online by the end of 2026, with another 10 GW potentially coming online in 2027. Musk also plans to have a 440,000-GPU Vera Rubin training cluster operational by the end of Q2 2027. To support the expansion, SpaceX is targeting an additional 20 GW of power and cooling infrastructure next year. Musk said there is a “non-zero possibility” of exceeding 20 GW of operational AI compute in 2027, although he acknowledged the probability remains small.

In just the first 72 hours since Gemini 4 Argon launched, we processed ~25 trillion tokens, the demand for intelligence is incredible and still under appreciated!

NVIDIA just released Flare, a local inference server built for DGX Spark and RTX Blackwell GPUs - Built for long chats, agents and tool calls - Serves from Rust, no Python - Plans memory up front - FP8 KV cache by default Interesting timing... right around when DeepSeek open-sourced their inference engine. Link below.

$SPCX just reported ABSOLUTELY INSANE earnings, and I don't think people realize how quickly Elon Musk is scaling this company. SpaceX generated a staggering $14.6 BILLION in quarterly revenue, representing 161% year over year growth, while earnings per share surged 197% to $0.35. That's a 12% revenue beat and a 119% EPS beat against the estimates shown. And the numbers get even CRAZIER. SpaceX reportedly generated $4.4B in net income and $8.8B in adjusted EBITDA, with EBITDA growing an incredible 306% year over year. We're talking about a company that's not only more than doubling its revenue but potentially expanding its profitability at an even faster pace.

China wants to control exports of its most powerful AI models. A new draft adds 23 AI technologies to its export control list, including large models, AI agents and computing clusters. Frontier models above 2 trillion parameters, or trained with more than 10^26 FLOPs, would need a license before their weights or source code leave the country. Open source models that are already public are exempt.

We are excited to work together with @FireworksAI_HQ on the early-launch of EAGLE-4. Check out our technical blog to understand how EAGLE-4 works and why it matters for creative workflows: https://t.co/YnYzQ89uCv https://t.co/02VMIubVgg

We built a robotics data engine with Astra-level performance at 125x lower cost. And we are open-sourcing the weights today! Introducing Flora, a suite of open models to parse the physical world and turn raw footage into rich training signals. Flora already powers our in-house data engine, labeling 200K hours of videos a day on 500 GPUs. We created it to scale the frontier for robot learning. Now you own the weights too. Flora gives you: - High-quality, timestamped captions to curate and filter human egocentric and robot data; - Dense progress rewards for quality control and RL post-training; - Rubrics scoring for real-world benchmarking Try it today, links in thread:

We're announcing six new features this week: Flex VMs, our own VM product that makes it significantly easier for us to onboard capacity. Spot, a mechanism for our suppliers and commitment users to monetize idle resources. CPUs and Apple Silicon, diversifying our compute resources for specialized use cases. MCP and CLI, making it seamless for humans and agents to access the world's compute. Learn more: https://t.co/ai73PLDbZY

🚀 Introducing DeepInfer, our in-house inference framework, now fully open source! Powering DeepSeek-V4.1-Flash at scale with high throughput and low latency. We’re sharing our production stack to help the community lower compute costs, reduce energy use, and make AI more affordable. 🔗 https://t.co/JGAik1kgk0 1/4

PyTorch 🤝 Google TPUs. TorchTPU is officially open source! 🚀

We're about to witness one of the BIGGEST wealth transfers in the history of the technology industry. According to Morgan Stanley, hyperscalers and neoclouds are projected to spend a staggering $1.58T in capital expenditures in 2027, including operating leases but what's even more interesting is where all of that money is actually going. We're talking about Google spending an estimated $375B, $AMZN $270B, $META $221B, $MSFT $209B, $SPCX $132B, $ORCL $97B, $CRWV $47B, and $NBIS $37B. These companies are collectively pouring more than a TRILLION dollars into building the infrastructure required to power AI, and that money doesn't simply disappear. It flows directly into the companies supplying the chips, networking equipment, memory, storage, power infrastructure, and data centers required to make all of this possible. Morgan Stanley estimates that approximately $672B will flow into semiconductors and servers, making this by FAR the largest category of spending. This is why companies like $NVDA and $AVGO remain so important to the AI buildout. Every time a hyperscaler or neocloud expands its compute capacity, it creates additional demand for GPUs, custom AI accelerators, and the supporting infrastructure. But it doesn't stop there. Another $215B is expected to flow into networking and optical infrastructure, benefiting companies that help connect these massive AI clusters together. As AI systems continue scaling, networking becomes increasingly important because having the fastest GPUs in the world means very little if they can't communicate efficiently. Then you have $217B going toward memory and storage which is why I remain so bullish on companies exposed to HBM, DRAM, NAND, and enterprise storage. And let's not forget the $119B flowing into data center infrastructure, including power equipment, cooling systems, electrical infrastructure, and the physical equipment needed to actually operate these facilities. And the way I see it, some of the biggest investment opportunities over the coming years won't necessarily come from the companies building AI models themselves, but from the companies supplying the critical infrastructure that everyone needs to compete. That's exactly why I remain so bullish on the picks and shovels side of this AI revolution and I believe we're STILL in the early innings of this entire infrastructure buildout.

We got our hands on @nvidia's new Vera CPU and put it to the test. We built a benchmark using real-world agentic workloads to see how it performs under contention. In early tests, Vera delivered up to 2x the throughput of comparable x86 hardware, maintaining low latency even as hundreds of concurrent workloads competed for resources. For Modal Sandboxes, that could mean running more agents in parallel without sacrificing latency.

Former $GOOGL employee on why data center capacity and electricity are the real bottlenecks, not GPU production itself, and how unreliable some of the neoclouds are ($AMZN, $GOOGL, $MSFT, $NVDA): 1. According to the expert, open-source models are closing the gap with frontier models faster than expected, with customers increasingly replacing Anthropic and OpenAI models in multi-agent architectures with open-source alternatives starting from Q2 of this year. The shift started with simple tasks like tool calling and MCP, but has moved up the stack to low-level orchestration. 2. The expert identifies capacity as the biggest constraint at the moment, with certain companies' traffic growing 5x while GPU availability, particularly in the U.S. where latency requirements are tightest, simply cannot keep up. Sourcing capacity overseas in Asia or Australia is technically possible but adds latency that most customers will not accept. 3. Hyperscalers offer reliability but are expensive, require multi-year commitments of three or more years, and have limited availability of the latest hardware, with customers already chasing B300s. AI-native companies tend to gravitate toward neoclouds due to hyperscaler pricing and contract rigidity, but neoclouds come with their own reliability issues, with GPUs dropping off unexpectedly and requiring engineering resources to manage fleet stability around the clock. 4. The third tier consists of smaller independent operators who have invested in GPU racks, seeing the opportunity to lease out capacity and recoup costs within a year or two, with the remainder being pure profit. The expert describes these as essentially unusable in practice, with operators who do nothing more than find land, install racks, connect electricity, and set up cooling, with no real cloud management, maintenance, or operational expertise behind them. 5. The expert expects overall GPU availability to increase meaningfully over the next year or two, with $NVDA continuing to ramp production. The real bottleneck is not GPUs themselves but data center capacity and electricity in the U.S. On the supply side, the expert also sees players $SPCX beginning to lease out spare capacity to the market, with more likely to follow. 6. The expert notes that U.S.-based neoclouds typically require 20-30% prepayment upfront, with the remainder paid monthly or quarterly over the contract term, and rebates negotiated at the end for any reliability incidents like traffic loss. APAC-based providers operate differently, with no prepayment required at all. found on @AlphaSenseInc

Former $GOOGL employee on why data center capacity and electricity are the real bottlenecks, not GPU production itself, and how unreliable some of the neoclouds are ($AMZN, $GOOGL, $MSFT, $NVDA): 1. According to the expert, open-source models are closing the gap with frontier models faster than expected, with customers increasingly replacing Anthropic and OpenAI models in multi-agent architectures with open-source alternatives starting from Q2 of this year. The shift started with simple tasks like tool calling and MCP, but has moved up the stack to low-level orchestration. 2. The expert identifies capacity as the biggest constraint at the moment, with certain companies' traffic growing 5x while GPU availability, particularly in the U.S. where latency requirements are tightest, simply cannot keep up. Sourcing capacity overseas in Asia or Australia is technically possible but adds latency that most customers will not accept. 3. Hyperscalers offer reliability but are expensive, require multi-year commitments of three or more years, and have limited availability of the latest hardware, with customers already chasing B300s. AI-native companies tend to gravitate toward neoclouds due to hyperscaler pricing and contract rigidity, but neoclouds come with their own reliability issues, with GPUs dropping off unexpectedly and requiring engineering resources to manage fleet stability around the clock. 4. The third tier consists of smaller independent operators who have invested in GPU racks, seeing the opportunity to lease out capacity and recoup costs within a year or two, with the remainder being pure profit. The expert describes these as essentially unusable in practice, with operators who do nothing more than find land, install racks, connect electricity, and set up cooling, with no real cloud management, maintenance, or operational expertise behind them. 5. The expert expects overall GPU availability to increase meaningfully over the next year or two, with $NVDA continuing to ramp production. The real bottleneck is not GPUs themselves but data center capacity and electricity in the U.S. On the supply side, the expert also sees players $SPCX beginning to lease out spare capacity to the market, with more likely to follow. 6. The expert notes that U.S.-based neoclouds typically require 20-30% prepayment upfront, with the remainder paid monthly or quarterly over the contract term, and rebates negotiated at the end for any reliability incidents like traffic loss. APAC-based providers operate differently, with no prepayment required at all. found on @AlphaSenseInc

WTF DeepSeek just gave away its production inference stack. Big move. DeepInfer, the engine behind DeepSeek-V4.1-Flash, is now open source, plus a component that makes community deployments match the official API. DeepSeek’s stated goal is lower compute costs and cheaper AI for everyone.

How could you POSSIBLY be bearish on compute right now when GPU rental prices are doing THIS? According to data from FastGPU, GPU rental prices increased as many as 81 times across providers over the past 60 days, with some of the most important chips used for AI training and inference seeing MASSIVE price increases. NVIDIA B300 rental prices increased 52.8%, going from $8.50 to $12.99 per GPU hour. H200 prices jumped 33.3%, B200 prices increased 27.8% and H100 prices rose 30% and the RTX PRO 6000? Rental prices surged an absolutely INSANE 307%. We're watching AI infrastructure companies get absolutely DESTROYED in the stock market, with investors freaking out about debt, financing costs, oversupply, and whether these companies will ever generate attractive returns on their investments. Meanwhile, the actual cost of renting the underlying compute is going UP. If we were genuinely heading toward a massive compute oversupply, you would expect rental prices to start collapsing as providers competed to fill their available capacity. Instead, we're seeing pricing strength across some of the most important GPUs powering the AI industry. And remember, we're still in the EARLY stages of AI adoption. Frontier labs are scaling their training clusters, inference demand is exploding as more people and businesses adopt AI, and the amount of compute required to serve increasingly sophisticated models continues to grow. This is EXACTLY why I've been so bullish on companies like $NBIS , $MU , $CRDO , $IREN etc. I genuinely believe compute will remain one of the most valuable resources in the AI economy for years to come. And while Wall Street continues panicking over the short term volatility, I'm BUYING THE DIP. If you want to know exactly what I’m buying on this dip and how I’m positioning across the compute buildout, check out portfolio below. https://t.co/Hkywss4mqJ

We're witnessing one of the BIGGEST supply demand imbalances in the history of the technology industry and it's showing NO signs of slowing down. According to a new report from TrendForce, hyperscalers are already locking in hard disk drive (HDD) supply for 2028, and even after securing long term agreements with manufacturers, they STILL expect shortages. HDD lead times are currently sitting at 50 weeks, compared to just 16 weeks in a balanced market. In other words, it can take nearly an entire YEAR to secure hardware that would normally take just a few months. TrendForce is reporting that DRAM lead times have reached 20 weeks compared to a normal 8 weeks, enterprise NAND storage is at 16 weeks versus 8 weeks, ABF substrates are sitting at an incredible 48–56 weeks compared to just 12 weeks and even MLCCs, which are critical components used in electronic circuits and power delivery, are experiencing significant supply constraints. Meanwhile, GPU lead times are actually relatively balanced. And I think this tells us something VERY important about where the AI buildout is heading. As AI inference scales, hyperscalers need enormous amounts of storage to support expanding datasets, model outputs, retrieval systems, and the infrastructure required to serve millions of users. And unlike GPUs, which handle computation, HDDs provide the cost effective, high capacity storage needed to retain enormous amounts of data. This is also why demand for enterprise SSDs is accelerating. When hyperscalers cannot secure enough HDD capacity, they increasingly explore high density QLC SSDs as an alternative, which can create additional pressure on NAND supply. And this is where things get REALLY interesting from an investment perspective. Companies like $WDC and $STX are positioned to benefit from constrained HDD supply, while $MU, $SNDK, and other memory manufacturers could benefit from increasing demand for DRAM and NAND. When customers are willing to secure capacity YEARS in advance, suppliers can gain stronger pricing power, better revenue visibility, and potentially higher margins. And that's why I continue to believe the AI infrastructure buildout is going to create enormous opportunities across the ENTIRE supply chain, not just GPUs. We're moving into an environment where having access to critical components, securing long term supply agreements, and actually executing on infrastructure deployment could become major competitive advantages. Of course, shortages don't automatically mean every supplier will see higher profits. Higher prices can eventually encourage supply expansion, customers may shift technologies, and elevated infrastructure costs could pressure demand but the BIGGEST takeaway for me is that hyperscalers are making purchasing commitments extending into 2028 while STILL anticipating shortages.

Billions in GB300s ... collecting dust in dark warehouses https://t.co/gfedEyPciR https://t.co/WbvK263gQA

Introducing SGLang Low-Latency Mode, built with @NVIDIAAI and @AMD. On DeepSeek-V4.1-Flash with DSpark, it delivers up to 1.38× P90 per-user token speed on NVIDIA GB300 (671 to 928 tok/s) and 1.56× on AMD MI355X (332 to 516 tok/s) in @SemiAnalysis_'s AgentX trace replay. Highlights: - NVIDIA: overlapping kernels cut decoder layer time from 96.8 to 68.4 µs - AMD: 2 persistent kernels per layer replace dozens of small ones, with cross-GPU communication overlapped inside them. - New plugin system: custom kernels plug in from external libraries, with automatic fallback. - Low-latency kernels for GLM-5.3, K3, and MiniMax M3 are in progress and will plug in next.

(1/6) The all-new FastVideo FastH3-QAD-v2 and the all-new H3 NVFP4 VAE decoders are here! 🚀🚀 This entire 768p video was generated on two DGX Sparks, using QAD-v2 and both new decoders. 8-step video + audio, QAD weights 💾 @NVIDIARTXSpark @NVIDIAAI @MiniMax_AI https://t.co/90G6RcD2xx

We trained our Muse foundational models, including Muse Wonder, at Prometheus, our gigawatt-scale supercluster in New Albany, Ohio. Across a supercluster with hundreds of thousands of GPUs, even one underperforming host can slow an entire training run. We engineered Prometheus for resilience at scale, achieving 85%+ goodput on pretraining, 90%+ on reinforcement learning and ~47K GPU hours mean time between interruptions during Muse Wonder training. Take a look at the tech powering Prometheus: https://t.co/WSFODnGRST

Yesterday, @OrnnExchange's B200 and B300 indexes reached a new all-time high prices, with the B200 hitting $8.54/GPU-hour. It is clear that today's market is lacking GPU capacity, and that the companies building deploying capacity are looking to improve their credit. In the past trading week as the B200 rental price increased by 12.9%, $CRWV fell by 25.2%. Why does $CRWV move inversely to rising compute prices as a neocloud? 98% of Coreweave's Q2 revenue came from committed contracts, 72% from three customers, and most of its $104B backlog is recognized after mid-2028. CoreWeave's backlog is locked at fixed prices. These higher market rates matter for new contracts they sign, so the question is where rental rates end up when their contracts roll off. Ornn's index tracks transacted prices in order to capture what buyers are actually willing to pay for on-demand GPU rentals, real signal for tracking the value of compute.

The AI infrastructure buildout is about to enter an ENTIRELY different phase in 2027. SemiAnalysis just released a new report forecasting 51.2 GW of new data center capacity in 2027, compared to 25 GW in 2026 and just 15.6 GW in 2025. We're talking about more than DOUBLING annual capacity additions in a single year. To put that into perspective, SemiAnalysis estimates that the industry could add roughly as much data center capacity in 2027 alone as it had built cumulatively through 2024, excluding China. And here's where things get REALLY interesting. While some investors, including Brad Gerstner, have questioned whether the industry can actually deliver this much capacity, SemiAnalysis points to satellite imagery showing more than 24 GW of data centers already in advanced stages of construction, with those facilities expected to be delivered in 2027. On top of that, the firm reports roughly $1 TRILLION in binding, non cancellable data center leases signed in 2026 so far. Now, think about what this means for the ENTIRE AI ecosystem. Every additional gigawatt of data center capacity creates enormous demand for GPUs, HBM memory, networking equipment, cooling systems, electrical infrastructure, and POWER. And remember, adding data center capacity doesn't automatically mean every facility will be fully equipped with GPUs on day one. In fact, SemiAnalysis is already raising the possibility that as power and construction constraints ease, the next major bottleneck could once again shift towards semiconductors and memory. And here's what I think a LOT of investors are missing. More compute capacity doesn't necessarily mean demand suddenly gets saturated. As inference becomes cheaper, AI adoption can accelerate across autonomous agents, enterprise applications, robotics, and eventually physical AI. Lower costs can unlock entirely new use cases that simply weren't economical before, creating another wave of demand for compute. If SemiAnalysis is even remotely close to being right, we're about to witness one of the largest infrastructure expansion cycles in history. And I believe the companies supplying the compute, memory, networking, and power behind this expansion could be some of the biggest beneficiaries over the next several years. If you want to see how I’m investing around this next phase of the AI infrastructure buildout, check out my full portfolio below: https://t.co/Hkywss4Ugh

Memory stocks just lost about $738 BILLION in market value. According to Goldman Sachs, shares of the major memory makers fell between 11% and 26% after China's CXMT announced its new DRAM cell architecture. SK Hynix fell about 25%. Samsung fell about 20%. Even CXMT itself dropped about 15%. Goldman still sees Samsung and SK Hynix moving to more advanced 1c nodes while CXMT stays mostly on older 1z technology through 2027. The market sold first. Now it has to decide if the tech gap really closed. $MU $SKHY

Rubin Ultra：Compute 送样两方案，材料同 Rubin 或树脂由 M8 升 M9 配一代布与 HVLP4，后者可能性较大但未定稿；可能 8 阶 30 层，也可能维持 5 阶 26 层；Switch 方案不清楚。 正交背板：前期用 M9 Q 布，因电性能不达标改考虑 PTFE，多层加工问题未解决、是否投线在观察；该板属 Kyber 机架、已推迟至 2028 年，Rubin Ultra 机柜不含正交背板。

What it’s like trying to rent AI/SI compute https://t.co/SBdEXK31zc

Morgan Stanley has revised AMD's GPU demand forecast for 2026-2030 from 14.1 million units to 8.0 million units, or -43.3%, while keeping Nvidia's the same. $AMD $NVDA https://t.co/U0nRij2XHd

NVIDIA ($NVDA) would incur an additional 60kW of power and $2M of cost per rack by replacing short copper cables with optical links, according to Bernstein's expert assessment with former NVIDIA leader Brad Smith. In the Vera Rubin architecture, NVLink switching sits centrally in the rack to connect compute trays above and below through roughly 5,000 short copper cables. These passive copper links operate across distances up to 18 inches at 200Gbps signaling speeds, retaining printed circuit boards with copper traces. Prior to accelerated computing workloads running across up to 1 million GPUs, copper handled roughly 90% of data center interconnects alongside 8% multi-mode optics and 2% single-mode optics, because electrical wiring provides unmatched capital and thermal efficiency. Physics dictates that achievable copper reach compresses as signaling speeds rise, falling from roughly 100m at 1Gbps to 2.5m at 100Gbps, 18 inches at 200Gbps, and 12 inches at 400Gbps. At 400Gbps, passive copper reach becomes impractical as electron radiation degrades signal integrity along the conductor. To bypass cabling congestion without abandoning electrical transmission, Feynman-era architectures approaching 1MW per rack integrate four blade assemblies directly into a midplane board with back-mounted NVLink switches. While active electrical cables with retimers can extend 200Gbps reach to 2 meters, they introduce roughly $1,200 of unit cost and 100 to 200 nanoseconds of latency. Server makers will therefore keep copper inside the rack until package-level input-output demands exceed short-reach electrical limits. #AdvancedPackaging #NVIDIA #TEConnectivity #OpticalNetworking #Networking Report date: 2026-10-15 · Bernstein

Anthropic says Claude will lead 80–90% of building the next Claude. The question for the rest of us is which 10% we keep. At the IMF AI & Cyber Risks panel, Anthropic's Michael Sellitto said Claude's share of leading next-gen model work went from 0% in January to 3% in April and 26% in August, and could hit 80–90% sometime in 2027. https://t.co/b2u0M9aTXj

Ben Horowitz on the lesson LoudCloud, his early cloud company, taught him about today's AI neoclouds, and why they now face the opposite problem: "It can end up being like a very dangerous arbitrage business, where you can get upside-down easily. And so if you look at LoudCloud, just as an example, here's a simple thing that happened to us." "In 1999, bandwidth was $1,500 a megabit. And you couldn't get it. It was very hard to get it. And so you had to get a long-term contract, three years. By 2001, it was $50 a megabit... And if you borrowed money to get it, you're dead." "That kind of thing is so out of the control of the entrepreneur." "I think the core dynamic of AI has been working for CoreWeave, though, in that there's a massive shortage, not an overage of supply... Token prices have gone down, but GPU prices have not." "And by the way, data center space is going up in price because even labor to build data centers is hard. The politics are making it hard to build data centers. So the whole supply chain price is going up and demand is not subsiding." "Those are very big tailwinds for CoreWeave, despite the fact that the debt markets, kind of independent of them, are tough." @bhorowitz w/ @chamath on @allinpod

Power is the BIGGEST constraint in the entire AI buildout right now. Morgan Stanley estimates that US data centers could require roughly 160 GW of power by 2029, but even after accounting for data centers under construction, additional utility grid access, behind the meter generation, and potential nuclear capacity, there could STILL be a 76.5 GW power shortfall. Think about how INSANE that is. We're talking about nearly HALF of projected US data center power demand potentially going unmet and here's why this matters so much. We can manufacture more GPUs, build better AI models, and raise hundreds of billions of dollars to finance new infrastructure. But none of that matters if you can't actually power the data centers. Electricity isn't something you can simply scale overnight. Building new transmission infrastructure, expanding grid capacity, securing interconnection agreements, and developing new power generation can take YEARS. And with AI workloads becoming increasingly compute intensive, demand for electricity is only going to accelerate as inference, autonomous agents, robotics, and physical AI continue scaling. This is exactly why I think POWER is becoming one of the most valuable assets in the entire AI ecosystem. And I think @elonmusk understands this better than almost anyone and it's also why I remain so bullish on neoclouds like $NBIS, $IREN, $SPCX. These companies are becoming incredibly important because they're helping solve one of the biggest bottlenecks in the AI industry turning massive demand for compute into actual operational capacity. The companies that can secure large amounts of power, develop data center campuses, deploy GPUs, and get that infrastructure operational FAST are going to have an enormous competitive advantage. And as power becomes increasingly scarce, the value of existing infrastructure with secured electricity and grid access could become significantly more valuable. You can't have an AI revolution without compute, and you can't have compute without POWER. If you want to see how I’m investing around the AI power bottleneck and the infrastructure companies I think benefit most, check out my portfolio below. https://t.co/MMkdjfqLsf

Huge thanks to @ParadisLabs, @ren_stocks, @pequityresearch, @StockSavvyShay, @zerohedge, and @jukan05, as many details were drawn from their work in the writing of this article. https://t.co/MWRfehTV4M

SHOCKING: Many neocloud executives we spoke with feel that if they have non-NVIDIA networking gear on their cluster, or if their cloud has an AMD GPU or TPU offering, NVIDIA will retaliate. They feel that retaliation includes not giving early allocation or no longer supporting a potential IPO/VC raise.(1/3)🧵

The amount of money Big Tech is about to pour into AI infrastructure is actually INSANE. Capital expenditures from the five major hyperscalers, $AMZN, $MSFT, $GOOGL, $META, and $ORCL are projected to increase from roughly $430 billion in 2025 to nearly $1.2 TRILLION annually by 2028. That's almost THREE TIMES as much spending in just three years. These companies are essentially rebuilding the computing infrastructure that the global economy will depend on for decades. Think about everything that goes into an AI data center. You need GPUs to train and run models, CPUs to manage workloads, HBM and DRAM to feed those processors with data, networking equipment to connect thousands of chips together, optical transceivers to move information at enormous speeds, and massive amounts of electricity to power and cool everything. And as AI models become more capable and adoption increases, the infrastructure required to support all of this becomes even more demanding. This is exactly why I've been so bullish on the companies supplying the infrastructure behind AI rather than trying to predict which AI model or application will ultimately win. Now, obviously, not every dollar of this spending will turn into profits for suppliers. Competition, pricing pressure and execution will matter a LOT, especially as these companies continue expanding capacity. But when you're looking at more than $1 trillion in annual spending from just FIVE companies, I think it's pretty clear why the opportunities across memory, networking, optical components, power, and semiconductors could be enormous. That's why I continue to believe AI infrastructure will be one of the biggest investment opportunities of this decade. The spending is already happening and the companies supplying the critical components are in a position to benefit for years to come.

🛰️ @elonmusk wants 10 gigawatts of compute by the end of 2027. @JensenHuang says Nvidia now counts @SpaceX alongside its cloud provider customers, and the two companies' engineers are designing a server for space: "In the two parts of our chart, Elon used to be under enterprise. That's the way we reported the numbers. Now he's part of our CSPs... They're massive. He's using our technology for his own AI in training models and such, deploying models, of course... And the demand for GrokBot, the consumption for GrokBot is really skyrocketing. And then, of course, he rents it. And so the combination of these three things, he's fairly unique in doing that... He loves the intersection between computing and the physical build-outs... And nobody knows more about energy than he does. And one particular new market that we're going to go explore together is putting data centers out in space. So our two engineering teams are working super closely together to create a new type of server, a new type of architecture that might be super good... in space and on Earth." — @JensenHuang, CEO of @nvidia on @BG2Pod with @altcap

Jensen Huang once admitted that when NVIDIA unveiled its first AI supercomputer, nobody wanted it. Not a single purchase order. Except for Elon Musk. "He goes, 'You know what? I have a company that could really use this.'" But then came the twist: it was a nonprofit😂😂 The machine cost $300,000, and Jensen thought the odds of getting paid were "approximately zero." He delivered it anyway, driving it to San Francisco himself in 2016 to a room full of researchers who turned out to be OpenAI. Elon saw the value of AI computing when the whole world couldn't understand it. He's done it again and again, spotting the future years before everyone else, then backing it with action instead of words.

THE STACK: ROBOTAXIS, BRAIN CHIPS & THE FUTURE OF AI Is @Tesla pulling away from the competition? Can @SpaceX keep breaking launch records? And is @Neuralink bringing science fiction to life? John Stringer (@teslaownersSV) and Joseph Jefferson (@SjvTesla) dive into @elonmusk's latest breakthroughs, the Robotaxi race, @grok's AI ambitions, and @Apple's reported partnership with @LGUS. How far can this tech revolution go? 0:11 - Tesla's massive Q3 numbers 4:39 - Tesla Robotaxi vs. Waymo 8:02 - SpaceX's rocket launch blitz 11:06 - Neuralink's brain-computer breakthrough 14:00 - X, Grok, and the AI race 18:00 - Is Apple falling behind? 19:17 - Apple's smart home partnership with LG 20:32 - What's next in tech?

Nathan Lambert of @trillium_labs says Chinese AI researchers' jaws drop to the floor when he tells them how much compute OpenAI has: "The Chinese labs are very technically excellent. They are more compute-starved." "Without a doubt, they have a comparable talent pool to draw on... To say that China is anything but equal on talent, you're gonna have a hard time because they just have so many people being trained at such a high level." "What they don't have is compute." "You talk to people there, and you're like, 'Oh, I heard an average OpenAI researcher has five thousand H100 equivalents that they can access at any time,' and like their jaw just falls to the floor. They're just like, 'I can't comprehend.'" @natolambert @lishali88

Everyone needs to STOP worrying so much about this neocloud selloff. I think people are forgetting just how important companies like $CRWV and $NBIS are to the entire AI ecosystem, especially $NVDA. And remember what Jensen Huang announced just TWO months ago? Nvidia partnered with BlackRock, Blackstone, Goldman Sachs, Apollo, Brookfield, and KKR to help mobilize over $500 BILLION in financing for AI infrastructure. Think about what that actually means. Jensen understands that the biggest bottleneck in AI isn't necessarily demand anymore, it's getting enough capital, power, and infrastructure to actually meet that demand. Also, Jensen NEEDS these neoclouds to succeed because they give Nvidia leverage against the hyperscalers. Companies like $MSFT, $AMZN, and $GOOGL are developing their own custom AI chips to reduce their dependence on Nvidia over time. And that's exactly why neoclouds are so important. Companies like CoreWeave and Nebius are building enormous amounts of GPU infrastructure centered around Nvidia's technology. They give AI companies another place to access compute without having to rely entirely on the traditional hyperscalers. The stronger these neoclouds become, the more alternatives customers have and the less dependent Nvidia becomes on a handful of hyperscalers controlling AI infrastructure spending. Now imagine what happens if these neoclouds suddenly struggle to finance their expansion, which is exactly what we're starting to see with $CRWV. CoreWeave's bond yields just surged to over 15%, which is a warning sign. For context, these bonds were issued at a 9.625% yield just four months ago and now investors are demanding significantly higher returns to hold that debt. Essentially, the bond market is telling us that investors are becoming increasingly concerned about CoreWeave's financial risks, which could make future borrowing a lot more expensive. And that's a massive problem for $NVDA because these companies are buying billions of dollars worth of GPUs and helping expand the market for Nvidia's technology. This is why I believe Jensen will step in to support strategically important partners like CoreWeave if financing conditions become difficult. Nvidia has already demonstrated that it's willing to put capital behind this ecosystem. I'm not saying Jensen will bail out every neocloud, but I find it very hard to believe he'll just sit back and watch one of his most important customers struggle to finance its expansion, especially when Nvidia needs these companies to compete against the hyperscalers.

NVIDIA has just announced a $250m compute commitment into the the UK! 🇬🇧 The money is to: “strengthen the UK’s agentic safety and security research capability” This includes $50m in accelerated computing access annually through BOLD at the University of Oxford, SOFAIR at UCL and Imperial College London.  NICE

Revenge of the GPU poor roboticists (MIT)

This is EXACTLY why I remain so bullish on the broader AI infrastructure buildout. Look at what is happening with GPU rental prices compared to AI token prices. Since July, the cost of AI tokens has collapsed, yet H100 GPU rental prices have remained elevated at roughly $2.70 per hour and this is a perfect example of Jevons Paradox. As AI becomes cheaper to use, developers and businesses can afford to use significantly more of it. Applications that were previously too expensive to run at scale suddenly become economically viable. Think about it, if the cost of running an AI agent drops significantly, companies can afford to deploy thousands of agents instead of hundreds. The same applies to coding assistants, AI search, customer support, and pretty much every application that relies on inference. And while each individual task might become cheaper, the total number of tasks being performed could grow significantly faster. That means more demand for GPUs, memory, networking equipment, data centers, and electricity and we're already seeing signs of this in the chart. Even as token prices have fallen sharply, H100 rental prices have remained relatively stable since July. That suggests demand for compute is still strong enough to support pricing despite the improvements we're seeing in AI efficiency. Now, there is an important distinction here. Cheaper tokens don't automatically guarantee higher GPU prices. Hardware supply, utilization rates, and newer GPU generations all play a role. But the fact that compute pricing has remained this resilient while token prices have declined tells us something important about underlying demand. I also think investors are underestimating just how much additional AI usage lower prices could unlock over the next several years. We're still relatively early in the adoption cycle, and many businesses have barely started integrating AI into their daily operations. This is why I continue to favor the companies supplying the infrastructure behind AI rather than trying to predict which model provider will ultimately dominate. Regardless of whether OpenAI, Anthropic, Google, or open source models capture the most market share, someone still needs to provide the compute required to run all of these models. And as long as AI usage continues growing faster than efficiency improvements reduce compute requirements, the broader infrastructure buildout should continue benefiting. Bullish on the broader AI infrastructure buildout. If you want to see how I’m investing across compute, memory, networking, and power, check out my full Milk Road Pro portfolio below: https://t.co/thIhK9ZH4E