PJFP.com

Pursuit of Joy, Fulfillment, and Purpose

Tag: AWS Trainium

  • Why the Markets Are Pricing AI Wrong: Gavin Baker on the July 2026 Selloff, GPU Spot Prices, Memory LTAs, and Nvidia’s Credit Wrapper

    Gavin Baker of Atreides Management returned to Invest Like the Best with Patrick O’Shaughnessy days after one of the strangest months the AI trade has ever produced. AI and semiconductor names fell 40 to 60 percent in a straight line while, by Baker’s account, not a single quantitative metric on the ground deteriorated. He spent the week in Silicon Valley hunting for a bearish data point and came back with almost nothing except credit. This conversation is the result: a detailed argument that the market has mispriced the gap between contracted compute and spot compute, that open source is growing the infrastructure pie rather than shrinking it, and that the one risk actually worth fearing is political rather than financial.

    TLDW

    Gavin Baker describes July 2026 as “2022 packed into a single month,” a violent AI and semiconductor drawdown that happened while hyperscaler operating cash flow accelerated from roughly 28 percent growth to 32 percent, or closer to 35 percent adjusting for unusual legal charges. His core claim is that the installed base of GPU compute is locked into long-term contracts priced far below the current spot market, so as those contracts roll off, compute reprices higher, operating cash flow accelerates, and the buildout can be funded internally rather than with the debt that widening credit default swap spreads and a poorly received Meta bond have made look expensive. He walks through each catalyst of the selloff: Meta renting out compute (misread as a capex cut), the open source capability leap from GLM 5.2 and Kimi K3 (misread as deflationary when a token is a token and costs the same flops, watts, and memory to produce), China acquiring a domestic deep ultraviolet lithography machine (real but 25 years behind), and rising real yields (the only genuine negative). He covers the game theory of breaking a memory long-term agreement in a world where market share is set by supply allocations, Nvidia’s new credit wrapper plus revenue share model and why it is misunderstood, the router and fine-tuning stack from Fireworks and Baseten that turns “ChatGPT wrappers” into defensible AI natives, continual learning as the one technical development that could disrupt training demand, SRAM accelerators for disaggregated inference, SpaceX as an underappreciated compute company with orbital ambitions, and his view that regulation, not fundamentals, is the biggest risk because the industry has done a terrible job telling its own story. He also makes an unusual observation about market structure: everyone now feeds news into Claude, and Claude has become a kind of Walter Cronkite for the stock market, collapsing the diversity of interpretation that normally keeps markets stable.

    Thoughts

    The load-bearing claim in this episode is the spread between contracted and spot compute, and to Baker’s credit it is falsifiable in a way most bull cases are not. He is not arguing that AI will be transformative or that demand feels strong. He is arguing something narrow and checkable: hyperscalers and neoclouds signed multi-year GPU contracts in 2024 and 2025 at prices that assumed a gentle decline, prices instead went vertical, and the installed base is therefore systematically under-earning. A startup rented several thousand B200s in the mid two dollars per GPU hour range and expects to pay just under four dollars for an identical cluster seven months later. If that repricing is real and broad, hyperscaler operating cash flow mechanically accelerates and roughly 700 billion dollars of projected credit demand evaporates. If GPU rental prices roll over and stay down for two consecutive quarters, the thesis is dead. That is the number to watch rather than any earnings headline. The caveat he steps past quickly is that the open source mix shift he describes as bullish does not eliminate margin, it relocates it, out of the frontier labs and down into the infrastructure layer. Excellent if you sell GPUs, power, and memory. Considerably more awkward for the labs whose projected cash flows are the reason anyone believes the compute gets paid for at all.

    The Claude as Walter Cronkite observation deserves more attention than it got, where it passed as a joke. Baker is describing a genuine change in market microstructure. Every institutional and retail participant now feeds the same news into roughly the same models, and while those models are probabilistic, they are not producing meaningfully diverse readings of the same headline. He connects this to Michael Mauboussin’s argument that a breakdown in diversity, not leverage alone, is what produces bubbles and crashes. If that is what happened in July, then the Japanese capacitor stock chart he cites, an entire three-year cycle compressed into six weeks before the fundamentals had even arrived, is not a curiosity. It is the signature of a market where thousands of participants share one interpretive engine. That makes drawdowns faster and deeper without making them more informative, which argues for holding through machine-generated narrative cascades rather than trading them.

    The middle of the conversation contains the most consequential business idea in it, and it is one that got almost no coverage during the selloff: memory long-term agreements and Nvidia’s credit wrapper are the same move executed at two different layers of the stack. Both trade near-term upside for durability. The memory companies stopped maximizing spot price and started signing prepaid agreements with floors and ceilings, and the reason those agreements will hold is that the penalty for breaking one has changed category. Apple could renege on memory pricing for years because its volume was overwhelming and it had no equivalent competitor. In a world with four buyers that matter and where AI market share is set by supply allocation rather than product quality, a supplier can answer a broken price agreement by breaking the volume commitment and handing your allocation to a rival, in an industry where oversupply is always followed by undersupply. Nvidia is running the same play one layer up. The credit wrapper with a revenue share above a price floor converts a cyclical one-time chip sale into a royalty on recurring compute revenue, financed on someone else’s balance sheet, which is a materially better business than selling hardware. It also widens the moat, because a startup accelerator pays more at the foundry, pays more for high bandwidth memory, and cannot finance its chips at Nvidia’s rate. Baker is right that this is misunderstood, and it is a strange thing for a stock at a ten-year-low forward multiple to be quietly doing.

    The technical material in the back half reveals an asymmetry worth naming. Baker treats two efficiency developments very differently. Continual learning and sample efficient learning, which several labs believe are close, would collapse the token budget required to produce a capable model, and he handles this by asserting that training asymptotes to a small but nonzero share of compute and that the outcome would be wonderful for the world anyway. SRAM-based accelerators for disaggregated inference, running prefill on one chip, attention on a high-memory chip, and the feed forward network on SRAM, he embraces enthusiastically as a return-on-investment improvement across the installed base. Both are efficiency gains. One is treated as neutral, the other as clearly positive, and Jevons paradox is doing all the work in both directions. That is probably correct given everything we have observed so far, but it is an assumption rather than a finding, and it is the assumption on which the entire “cheaper compute is bullish for compute” framework rests. Worth noting too that the SRAM disaggregation point is genuinely underdiscussed: those chips sit on older nodes and do not compete for leading-edge capacity, so they are additive supply rather than substitute supply.

    The final twenty minutes hold both the largest unpriced upside and the largest unpriced risk, and neither is in consensus estimates. On the upside, only the hyperscalers, CoreWeave, Crusoe, and SpaceX have ever brought more than 500 megawatts online in a single year, and SpaceX has done it fastest and cheapest. When it dumped a large block of compute into the market, the market absorbed it without a blip, which tells you more about demand than any survey. Baker’s sanity check on orbital compute is the sharpest reasoning move in the episode: Benchmark, from entirely outside the Elon ecosystem and without the benefit of internal launch costs, funded StarCloud at a real valuation, so the set of people who would all have to be wrong keeps growing. On the downside, regulation is the risk he names first and it is the one his own framework cannot arbitrage. New York’s data center moratorium is not a fundamentals problem, and no amount of operating cash flow acceleration fixes a permitting ban. His diagnosis is that the industry finds the benefits so obvious that it never learned to explain them, which is how a water usage figure overstated by four orders of magnitude became conventional wisdom. Proposing a foundation that buys World Series ad time is a tell about how far behind he thinks the industry is. Every other risk in this conversation is priced somewhere. That one is not.

    Key Takeaways

    • Baker characterizes July 2026 as “2022 in a month,” with AI names down 40 to 60 percent from their highs in a straight line while underlying fundamentals improved.
    • He spent the week in Silicon Valley explicitly hunting for a negative quantitative metric and found essentially one: third-party data suggesting Anthropic’s growth curve came slightly off trajectory, a data point Anthropic shareholders reportedly dispute.
    • Nvidia was trading at its lowest forward price to earnings multiple in ten years at the time of recording. The only cheaper moments were the DeepSeek shock and Liberation Day, both of which proved to be V-bottoms.
    • A low forward multiple means the market believes these companies are significantly over-earning. Baker’s counter is that they are under-earning because their installed compute is contracted below spot.
    • Combined operating cash flow at Microsoft, Meta, and Amazon accelerated from roughly 28 percent to 32 percent growth, or to about 35 percent after adjusting for an unusual quarter of legal and regulatory charges.
    • Nobody in 2024 or 2025 modeled old GPU prices going vertical in 2026. The bull case assumed a slow decline in rental rates and the bear case assumed a steep one.
    • A concrete example: a well-known startup rented several thousand Blackwell B200s in the mid two dollars per GPU hour range and expects to pay just under four dollars for an identical cluster seven months later, a 50 to 60 percent increase.
    • One inference cloud stated publicly that it plans to pay roughly 100 percent more for Blackwells when its current contract expires.
    • Neoclouds were often forced into below-market long-term contracts because they needed an offtake agreement to finance the GPUs in the first place.
    • Consensus models hyperscalers monetizing Blackwell and Rubin at roughly Ampere rates, two generations behind, producing about 1.3 to 1.4 trillion dollars of hyperscale operating cash flow. Assuming monetization merely at a discount to current Blackwell rates pushes that closer to two trillion and removes roughly 700 billion dollars of credit demand.
    • The credit concerns are real and undeniable: real yields are up, spreads have widened, credit default swap levels for the large buyers have blown out, and a recent Meta bond did not price where a Meta bond should price.
    • Baker’s response is that debt-fueled buildouts demand immediate repayment and unwind violently, which is what happened in the internet buildout, but this buildout is still overwhelmingly funded from operating cash flow.
    • If credit is not available, he argues the existing flops simply become more valuable, which is self-correcting rather than catastrophic.
    • The Meta selloff catalyst was a misread. Meta renting out compute was interpreted as excess capacity and a capex cut. Meta did not cut capex, and the actual motivation appears to have been demonstrating strong internal rates of return on a small slice of capacity ahead of a capital raise.
    • The open source panic was also a misread. Open source taking token share moves margin dollars out of the frontier model layer, but a token still requires the same flops, memory, and watts to produce, so infrastructure demand rises rather than falls.
    • Frontier tokens carry gross margins somewhere in the 80 to 95 percent range. Open source tokens might carry 30 percent. The customer’s savings come almost entirely out of that margin, not out of compute consumption.
    • Baker calls open source “dark matter to the public markets,” growing rapidly through GLM 5.2, Kimi K3, and Nvidia’s Nemotron, but nearly impossible for public investors to measure since it runs through private inference clouds.
    • Jensen Huang being the world’s loudest supporter of open source is itself evidence that open source is good for Nvidia’s business.
    • Enterprises that blow through their AI budget in three months set up a router, which cuts their spend but often increases total GPU hours consumed by shifting volume to cheaper open source tokens.
    • Adoption is happening in staggered waves: AI natives are all in and hiring very few humans, coastal public companies are optimizing, East Coast and non-coastal companies have barely adopted, and Europe is trying to regulate AI before using it.
    • Roughly 500,000 people worldwide use agentic AI, and perhaps half that number use it seriously, yet the world is already in an acute compute shortage. The relevant question is what happens at 100 million or 500 million users.
    • Token spend at the most AI-forward companies now runs 20 to 25 percent of total compensation spend, with individual examples at 30 percent and reports as high as 50 percent, against a roughly 25 trillion dollar global knowledge work market.
    • Founder-controlled companies are not conducting large-scale layoffs, which suggests the cash flow to pay for AI is expected to come from growth rather than from labor substitution.
    • Memory is the dominant variable in token economics. More memory per unit of compute yields more tokens out, which lowers cost per token, which is why demand has shown no negative elasticity to memory pricing.
    • Memory suppliers have shifted from maximizing near-term price to signing long-term agreements with prepayments, floors, and ceilings, trading short-term upside for durability.
    • Breaking a memory long-term agreement is now potentially fatal. With four buyers that matter at scale and market share determined by supply allocation, a supplier can respond by breaking the volume commitment and handing your allocation to a competitor.
    • This is structurally different from the Apple era, when a single dominant buyer could break pricing agreements without consequence.
    • Nvidia’s new model is best described as a credit wrapper with a revenue share triggered when GPU prices exceed a floor. It is not vendor financing, since a third party lends the money, and it could produce a very large cloud-scale royalty business quickly.
    • Baker thinks this model is badly misunderstood, meaningfully increases Nvidia’s revenue per gigawatt, and strengthens its competitive position against startup accelerators that pay more at the foundry, pay more for high bandwidth memory, and cannot finance their chips as cheaply.
    • Nvidia has taken equity stakes across the ecosystem, and Baker’s read is that every time they have not taken a stake it has proven to be a mistake.
    • The scenario that would genuinely frighten him: hyperscaler operating cash flow stops accelerating, forcing the buildout onto debt, or a sustained sharp contraction in GPU rental prices. Nobody he has spoken to says they have too many GPUs.
    • Continual learning and sample efficient learning are the technical developments most likely to disrupt training demand, and several new labs including Safe Superintelligence are focused on them. Baker still thinks training asymptotes to a small share of compute rather than to zero, and that the change would be enormously good for the world regardless.
    • Fireworks launched a product called Nexus that plugs into Claude Code, OpenAI Codex, or Grok in roughly three lines of code, ingests a customer’s data, applies reinforcement learning to a model, and routes queries appropriately.
    • This stack is what converts an alleged “ChatGPT wrapper” into a defensible company. Shifting 30 to 60 percent of token consumption to a customized open model on top of frontier orchestration produces better outcomes at roughly half the cost.
    • Cheap, capable open source models may actually inflate the value of the very best frontier model, since a 160 IQ orchestrator becomes more valuable when it has an army of cheap 120 IQ models to direct.
    • The inference clouds are growing almost as fast as the frontier labs did in their early days while burning very little cash, which is extraordinary by any conventional software metric.
    • China obtaining a domestic deep ultraviolet lithography machine is a genuine phase transition and should not be dismissed, but the technology is roughly 25 years behind extreme ultraviolet, and lithography progress is learning by doing that cannot be teleported through.
    • Baker considers regulation the biggest single risk to AI, citing New York’s data center moratorium as the first of many and describing the current environment as post-factual and post-logical.
    • The public narrative that data centers raise power bills, drain water, and destroy jobs is largely wrong. Behind the meter deals typically lower local electricity prices, and modern community agreements include hospitals, schools, police and fire stations.
    • The widely cited data center water figure originated in a published error overstating usage by roughly 10,000 times, since acknowledged by the author, which Baker likens to the decimal point error that created the myth that spinach is exceptionally high in iron.
    • He argues data centers are among the best things to happen to blue collar wages in his lifetime, with ongoing rather than one-time employment from maintenance, replacement, and upgrade cycles.
    • SRAM-based accelerators built on older nodes and free of high bandwidth memory constraints could substantially improve return on investment by allowing disaggregated inference: prefill on one chip, attention on a high-memory chip, and the feed forward network on SRAM.
    • SpaceX has improved fundamentally since going public, and Baker believes the market does not yet understand it as a compute company. Only the hyperscalers, CoreWeave, Crusoe, and SpaceX have ever brought on more than 500 megawatts of power in a single year, and SpaceX has done it fastest and cheapest.
    • A widely circulated report claims SpaceX intends to bring on eight gigawatts of compute in 18 months. Baker doubts the number but notes that at roughly 50 billion dollars of monetization per gigawatt, even a fraction of it dwarfs the current consensus estimate.
    • When SpaceX dumped a large block of compute into the market, it was absorbed without a blip, which Baker reads as one of the more bullish demand signals of the year.
    • Orbital compute feels more real every day. Benchmark funding StarCloud, from outside the Elon ecosystem and without access to internal launch costs, functions as a useful sanity check on the idea.
    • Dark horse names Baker flags for the next phase: Lip-Bu Tan, Lin Qiao at Fireworks, and Scott Wu at Cognition.

    Detailed Summary

    A Selloff That Contradicted Every Fundamental

    Baker opens by describing July 2026 as 2022 compressed into a single month. AI names fell 40 to 60 percent from their highs in a nearly straight line. What made the month unusual was not the magnitude but the absence of a legible cause. In 2022 the market feared recession, rising rates, and inflation. During the DeepSeek shock and Liberation Day you knew exactly what the market was reacting to. This time the fundamentals moved in the opposite direction from the tape. GPU availability tightened, GPU rental pricing rose, DRAM spot prices rose, and token growth accelerated. Baker asked Patrick, who had also spent the summer in Silicon Valley, whether he had heard a single negative quantitative metric or a single instance of deceleration. The answer was nothing.

    Part of the problem is visibility. Public markets cannot see Anthropic or OpenAI directly, and they cannot see the American open source inference clouds like Fireworks, Baseten, Modal, and Together that monetize inference. Everyone stares at the same chart of semiconductor cash flow rising while hyperscaler free cash flow falls, and that chart omits the private companies entirely. It also omits the repricing dynamic Baker considers the most important fact in the market.

    The Spot Versus Contract Gap

    In 2024 and 2025 every serious forecast assumed GPU rental prices would decline, with the only debate being how fast. Neoclouds locked in long-term contracts partly out of prudence and partly because they needed offtake agreements to finance the hardware at all. The result is a large installed base of contracted compute trading at a steep discount to today’s spot market. Baker’s argument is that as those contracts roll off, compute reprices higher even if spot itself declines from current levels, and that repricing flows directly into hyperscaler operating cash flow.

    The anecdotes are stark. A prominent startup rented several thousand B200s in the mid two dollar per GPU hour range and expects to pay just under four dollars for an identical cluster seven months later. One inference cloud said publicly it plans to pay roughly double for Blackwells at contract renewal. Baker’s read is that hyperscalers are therefore under-earning across the board, which is the exact opposite of what a ten-year-low forward multiple implies the market believes.

    Financing the Buildout and the Credit Question

    Credit is the one bearish input Baker concedes is real. Real yields have risen, spreads have widened, credit default swap levels have blown out across the large buyers, and a recent Meta bond did not price the way a Meta bond should. Sophisticated private capital investors told him this is just banks hedging commitments, but he acknowledges the optics are bad and the facts are undeniable. His concern is the classic capital cycle: debt-financed buildouts demand immediate repayment, so when supply and demand slip out of alignment the unwind is fast and brutal, exactly as it was in the internet buildout.

    The math he ran is the counterweight. Consensus effectively models hyperscalers monetizing Blackwell and Rubin at Ampere rates, two generations behind, producing 1.3 to 1.4 trillion dollars of operating cash flow. Assume instead that they monetize merely at a modest discount to current Blackwell rates and the figure approaches two trillion, taking about 700 billion dollars of credit demand off the table. Better cash flow also improves the credit ratios, which makes debt cheaper if they choose to use it. And if credit disappears entirely, the flops already installed simply become more valuable. Microsoft brought on a large slug of capacity in June that did not even appear in second quarter results.

    How the Month Actually Unfolded

    Baker walks the sequence of catalysts. First, Meta announced it would rent out compute, which the market read as excess capacity and an imminent capex cut. Meta did not cut capex. What Meta appears to have seen was SpaceX selling trading-optimized clusters into the market at an enormous premium to contracted rates, and the plan was likely to demonstrate strong returns on a small slice of capacity before raising equity capital and increasing capex. Shortly afterward Meta released its best model in a long time, overshadowed by a competing release but a clear signal it was not easing off.

    Next came the open source freakout. Kimi K3 arrived, the widely watched token index dipped and flattened, and the two were connected: the index captures mix, and a shift from expensive frontier tokens toward open source tokens looks like weakness even when total compute consumption is rising. Then China’s deep ultraviolet lithography news triggered a broad selloff in semicap equipment. Finally, rising real yields and widening spreads gave the market a genuine reason to worry. Baker’s summary is that with the sole exception of credit, every one of these narratives was factually wrong, and a friend at Fidelity described the winning strategy of the past three years as doing the dumbest, most superficial thing as fast as possible and cycling between them.

    Open Source as Dark Matter

    The most important conceptual argument in the episode is that a token is a token. Regardless of which model produces it, a token consumes the same flops, the same memory, and the same watts. Open source taking share therefore does not reduce compute demand. It transfers margin from the frontier model layer, where gross margins might be 90 percent, to open weights inference at perhaps 30 percent, and the resulting price decline drives elasticity in token volume. Since frontier labs and open source models both run on the same underlying cloud infrastructure at the same compute cost, the effect is to push margin dollars down into the infrastructure layer.

    Baker calls open source dark matter to public markets. It is real, it is accelerating on the back of capability leaps from GLM 5.2 and Kimi K3, Nvidia continues to push Nemotron closer to the frontier, and yet none of it appears in audited financials that public investors can underwrite. He also notes the tell that should have settled the debate: Jensen Huang is the world’s most vocal supporter of open source, which would be an odd position for the largest beneficiary of frontier concentration to hold if open source actually threatened the business. Baker adds a normative point, that a world with only one or two dominant frontier models charging 90 percent margins is not good for humanity, and that many models is the better outcome.

    Routers, Fine-Tuning, and the End of the Wrapper Insult

    The practical mechanism behind the open source surge is the router plus fine-tuning stack. Inference clouds have become genuinely good at supervised fine-tuning and reinforcement learning, so a company can take its proprietary data, customize an open weights model, put it behind a router, and have the router send most queries to that model while escalating to a frontier model for verification or harder work. The result is often slightly better outcomes at half the cost. Fireworks shipped a product called Nexus that connects to Claude Code, OpenAI Codex, or Grok in roughly three lines of code and handles ingestion, reinforcement learning, and routing.

    This changes the durability question for AI natives. Two years ago the criticism was that these companies were thin wrappers with no defensibility. Now a company with domain-specific proprietary data can train on it, own the model serving 30 to 60 percent of its tokens, and get off the frontier lab treadmill it previously had no choice but to accept. Baker points to Cursor, Harvey, and others leaning hard into this. He also raises the counterargument fairly: some believe that once a frontier model achieves recursive self-improvement it will serve every intelligence level more cheaply through distillation, leaving no room for open source. He does not dismiss it, but he thinks the proprietary data held by AI natives and the orchestration value of the single smartest model make the multi-model future more likely. Cheap 120 IQ models arguably make a 160 IQ orchestrator more valuable, not less.

    Where the Money Comes From

    The pushback Baker gets on X is fair: even if hyperscalers are under-earning, where does the customer revenue ultimately come from? Definitionally it must come from faster economic growth through productivity or from labor substitution. He sees labor substitution happening at AI natives, though not through firing. They simply never hire the humans, and gross profit dollars per full-time employee at these companies is vertical compared with prior startup generations. Token spend now runs 20 to 25 percent of total compensation spend at the most aggressive companies, with individual examples at 30 percent and reports as high as 50 percent, against a roughly 25 trillion dollar global knowledge work market.

    The encouraging signal is that founder-controlled companies, the ones most likely to move fast on efficiency, are not conducting large-scale layoffs once you adjust for pandemic-era overhiring. That suggests they see continued opportunity for people plus large token budgets rather than a straight substitution. Data from Cognition, Ramp, and Stripe indicates that companies spending the most on AI are growing meaningfully faster, though Baker acknowledges the skeptics’ point that these datasets do not control for industry.

    The Memory Supply War and LTA Game Theory

    Everything is currently in shortage, and Baker argues the constraint is energizing gigawatts rather than manufacturing. Turbine makers and diesel generator makers are ramping, old aircraft turbines are being stripped and reconditioned for data center power, and regulatory policy is moving favorably. The transition he says he got wrong is the shift, especially in memory, from maximizing short-term pricing to signing long-term agreements with customer prepayments, price floors, and price ceilings.

    The reason those agreements will hold is game theory. Memory is the axis around which everything else revolves, because more memory per unit of compute means more tokens out, which lowers cost per token, which is why demand has shown essentially no negative elasticity. Market share among the four buyers that matter (Amazon with Trainium, Google with TPUs, AMD, and an Nvidia bigger than all of them combined) will be determined for years by supply chain allocation. Break a long-term agreement to chase a lower price in an oversupply year and the supplier can break the volume commitment in return and hand your allocation to a competitor. Since oversupply in this industry is reliably followed by undersupply, that is a decision that can end a franchise. Apple could get away with this historically because its volume was overwhelming and it had no equivalent competitor. That world is gone.

    Nvidia’s New Playbook

    Baker finds Nvidia’s low multiple hard to reconcile with how thoroughly the current environment favors it. If chips need to be financed, nothing on earth is more financeable than an Nvidia GPU. If land and power are the constraint, Nvidia has been playing the matchmaking chess game well. On top of that they have rolled out what Baker describes as a credit wrapper with a revenue share that kicks in when GPU prices sit above a floor. It is not vendor financing, since someone else lends the buyer the money. What it does is give Nvidia a royalty on recurring compute revenue, which could amount to a very large cloud business built entirely out of royalties, while helping bridge the cash flow mismatch between an industry that has gone free cash flow negative and a supplier collecting all the cash.

    Asked what he would do as a memory CEO, Baker says he would do exactly what Nvidia is doing: approach GPU and accelerator buyers, participate in the credit wrapper, perhaps put up cash upfront to make lenders comfortable, and take a cut of ongoing revenue. He expects firms like Blackstone and Apollo are pitching variants of this to the memory companies already. He also thinks the arrangement quietly widens Nvidia’s competitive moat, since startup accelerator companies pay more at the foundry, pay more for high bandwidth memory, and cannot finance their chips at Nvidia’s rate. And he notes that essentially every time Nvidia has declined to take an equity stake in something, it has turned out to be a mistake.

    What Could Break the Thesis

    Pressed for the scenario that would flip him, Baker names two. The first is operating cash flow failing to accelerate, which would force the buildout onto debt and validate the credit bears. That outcome depends largely on whether the combined trajectory of Anthropic, OpenAI, Grok, Cursor, and open source keeps compounding. The second is a sustained sharp contraction in GPU rental prices. The market would react instantly, and it would mean the compute shortage had broken. As of the recording, not a single person he has spoken with says they have too many GPUs.

    The technical wildcard is continual learning and sample efficient learning. Many researchers believe both are close. A human learns effectively on something like 20 billion tokens while frontier models train on 300 trillion, so a model that could be trained on 10 trillion tokens and then learn efficiently in the world would represent a discontinuity in training demand. Baker thinks training will asymptote to a small but nonzero share of compute regardless, and that the development would be extraordinarily good for the world. He also notes Nvidia is deeply involved with essentially all of the labs pursuing it.

    China, Lithography, and Decoupling

    On China’s deep ultraviolet lithography machine, Baker holds both views at once. It is a genuine phase transition, comparable to going from having no propeller plane to having one, because they did not have it before and now allegedly they do. It is also roughly 25 years behind extreme ultraviolet, and lithography is learning by doing, so you cannot teleport through the required cycles. He suspects the market overreacted and that if it ever affects ASML’s order book it will be years out, by which time the market will have forgotten and rediscovered the concern several times.

    He is careful about certainty here. It is very hard for an American to have real clarity on what is happening inside China, the people there are extremely capable and work brutally hard, and they consider this existential for the country. There are unverified reports that an extreme ultraviolet machine was smuggled in, which he treats as noise. His larger point is that decoupling is now self-reinforcing on both sides, it is unfortunate, and neither side is going to stop.

    Regulation, Data Centers, and a Failure of Storytelling

    Asked for the worst thing that could happen to AI, Baker answers regulation without hesitation. New York’s data center moratorium feels like the first of many, and even deep red pro-growth states are telling the industry it is doing a poor job explaining itself. The political narrative among ordinary Americans is that data centers will raise electricity prices, drain water supplies, and eliminate jobs. Baker’s counter is that behind the meter deals generally lower local electricity prices, that community agreements now routinely include hospitals, schools, police stations, and fire stations rather than the old model of buying the fire department new trucks, and that the jobs are ongoing rather than one-time because of continuous maintenance, replacement, and upgrade cycles.

    The water claim is the clearest case of a myth outrunning the correction. An author overstated data center water usage by roughly 10,000 times, has acknowledged the error repeatedly, and the figure still circulates. Patrick offers the parallel of the spinach iron myth, created by a misplaced decimal point in an academic text and still believed 80 years later. Baker’s proposed remedy is blunt: a foundation or political action committee running ads during the Final Four, NFL games, and the World Series explaining what a data center actually does for a community, alongside the story of AI accelerating medical research and improving outcomes for people with serious illness. The people building this find the benefits so obvious that they assume everyone already knows, and they cannot process how divergent their view is from most Americans.

    SRAM Accelerators and Disaggregated Inference

    An underdiscussed development, Baker argues, is what happens when SRAM-based accelerators arrive at scale. These chips are not constrained by high bandwidth memory and are often built on older nodes, so they do not compete for the leading edge capacity that GPUs consume. Inference disaggregates into prefill and decode, and decode splits further into attention and the feed forward network. The holy grail is running prefill on a chip without high bandwidth memory, attention on a high-memory chip, and the feed forward network on SRAM, which nothing beats for that workload. Since workloads keep changing, no single chip can get the ratio of compute to high bandwidth memory to on-die SRAM permanently right, which is precisely the argument for disaggregation. Baker expects this to be strongly positive for the return on investment across the installed base and on new compute.

    SpaceX, Orbital Compute, and Dark Horses

    Baker does not think the market understands SpaceX as a company yet, and he considers it the most important new public company. The fundamentals have improved since the IPO, and the compute story is the part being missed. Only the hyperscalers, CoreWeave, Crusoe, and SpaceX have ever brought more than 500 megawatts of power online in a single year, and SpaceX has done it fastest and cheapest while building clusters customers actually like. When SpaceX dumped a large block of compute into the market, it was absorbed without a blip, which Baker treats as one of the most bullish demand datapoints available. A circulating Substack report claims eight gigawatts within 18 months. He doubts that figure and quotes it only because it is public, but at roughly 50 billion dollars of monetization per gigawatt against a 73 billion dollar consensus estimate, even partial delivery would overwhelm expectations. There is a well-known New York hedge fund short case built on spot compute prices falling 90 percent.

    On orbital compute, Baker says time at Starbase left him thinking it feels more real every day, and the Starship landing reinforced it. His sanity check is that Benchmark, from entirely outside the Elon ecosystem and without the benefit of internal launch costs, chose to fund StarCloud at a real valuation, with SpaceX partnering to provide the Starlink laser technology that orbital compute requires. As he puts it, maybe he is crazy, maybe Elon is crazy, maybe Benchmark is crazy, and maybe the SpaceX engineers are crazy too, but all of that being true simultaneously does not seem probable. Asked for dark horses who could become as consequential as the current giants, he names Lip-Bu Tan, Lin Qiao at Fireworks, and Scott Wu at Cognition. The episode was recorded at Benchmark’s offices, at the table where their dinners are held.

    Notable Quotes

    “I want to be scared. I don’t want to feel like a lunatic watching these stocks get cheaper thinking the expected forward returns are going up.”

    Gavin Baker, on why he spent the week in Silicon Valley hunting for bearish data

    “I would describe July as 2022 in a month.”

    Gavin Baker, characterizing a 40 to 60 percent drawdown in AI names that happened in a straight line

    “Have you heard a single negative quantitative metric about AI? A single instance of deceleration?”

    Gavin Baker to Patrick O’Shaughnessy, framing the central contradiction of the month

    “A token is a token, and you need the exact same amount of compute to make a token. It takes the same amount of flops, the same amount of memory, the same amount of watts.”

    Gavin Baker, on why the open source panic misread infrastructure demand

    “Open source is kind of dark matter to the public markets. It’s hard for public markets to measure it.”

    Gavin Baker, on why the fastest-growing part of inference demand is invisible in audited financials

    “Claude is kind of Walter Cronkite for the stock market and everybody just believes whatever it says. And by the way, it’s really smart, but it’s not always right.”

    Gavin Baker, on the collapse of interpretive diversity among investors

    “Nvidia is actually, as we record this, at its lowest forward PE of the last 10 years.”

    Gavin Baker, noting the only cheaper moments were the DeepSeek shock and Liberation Day, both V-bottoms

    “If you break your LTA and then in the next two or three years for any reason leverage shifts back to the memory guys, you’re out of business.”

    Gavin Baker, on why long-term agreements will hold through the next memory cycle

    “If you need to be able to finance the chips, and you do, nothing’s more financeable than an Nvidia GPU. Nothing.”

    Gavin Baker, on why the current environment favors Nvidia more than its multiple suggests

    “Data centers are in a lot of ways the best thing to happen for blue collar wages in my lifetime.”

    Gavin Baker, on the gap between the political narrative and the local economics

    “A lie could go around the world faster than truth gets out of bed.”

    Gavin Baker, on a data center water usage figure overstated by roughly 10,000 times that still circulates

    “One of Elon’s phrases is we specialize in making the impossible late.”

    Gavin Baker, on why he doubts the eight gigawatt figure without betting against SpaceX

    Watch the full conversation here: Why the Markets Are Pricing AI Wrong with Gavin Baker on Invest Like the Best.

    Related Reading

    • Invest Like the Best on Colossus the show’s home, where the full episode archive and transcripts live.
    • Atreides Management Gavin Baker’s firm and the vantage point behind these compute and semiconductor calls.
    • More Than You Know by Michael Mauboussin, the source of the diversity breakdown framework Baker invokes to explain why markets crash when everyone reasons the same way.
    • High Bandwidth Memory (Wikipedia) background on the memory technology that sits at the center of the long-term agreement game theory.
    • Fireworks AI the inference cloud whose routing and fine-tuning stack Baker credits with making open source models competitive for production workloads.

  • OpenAI and Broadcom Unveil Jalapeño, a Custom LLM Inference Chip to Cut Compute Costs and Reduce Nvidia Dependence

    OpenAI and Broadcom pulled the wrapper off Jalapeño on Wednesday, June 24, 2026, a custom silicon accelerator that OpenAI is calling its first “Intelligence Processor” and its first real move into designing the hardware underneath its own models. Broadcom President and CEO Hock Tan and President Charlie Kawwas physically handed the wafer to OpenAI CEO Sam Altman and President and Co-Founder Greg Brockman, a staged moment meant to signal that the ChatGPT maker is no longer just a models-and-products company but is now reaching all the way down to the chip. Jalapeño is purpose-built for large language model inference, the compute-intensive job of actually serving answers to users rather than training the model in the first place, and OpenAI plans to deploy it at gigawatt scale by the end of 2026 as the first step in a multi-generation platform built with Broadcom and Canadian electronics manufacturer Celestica. You can read the announcement straight from the source in OpenAI’s official post.

    TLDR

    OpenAI and Broadcom unveiled Jalapeño, OpenAI’s first custom AI chip, an ASIC designed from a blank slate specifically for LLM inference rather than training, manufactured by TSMC and integrated into server systems by Celestica that only OpenAI will use. OpenAI claims the chip went from initial design to manufacturing tape-out in just nine months, what it calls the fastest ASIC development cycle ever in high-performance advanced semiconductors, accelerated in part by using its own AI models to design the silicon. Engineering samples are already running ML workloads in the lab, including GPT-5.3-Codex-Spark, and OpenAI says early testing shows performance per watt “substantially better” than current state-of-the-art, a self-reported and not yet independently verified claim with a full technical report promised in the coming months. Broadcom CEO Hock Tan told Reuters the chip matches Nvidia’s Blackwell and Google’s TPUs, framing the launch as part of a flywheel where OpenAI owns the full stack from chip to model to product. The chip slots into a broader infrastructure strategy targeting 10 gigawatts of custom accelerator capacity between 2026 and 2029 with deployments alongside Microsoft and other partners, and The Decoder reported Microsoft is expected to buy 40 percent of the chips, a guarantee Broadcom reportedly demanded to secure the first phase. The move is widely read as OpenAI diversifying away from Nvidia, continuing a procurement spree that already includes AWS Trainium, AMD, and Cerebras, as inference quietly becomes the company’s real cost center.

    Thoughts

    The single most important word in this announcement is “inference,” and it is the word doing the heavy lifting. Training a frontier model is a capital expense that happens in bursts. Inference is the bill that arrives every single day, forever, scaling linearly with usage. Every ChatGPT reply, every Codex task, every API call, every agent step is an inference event, and as OpenAI’s product surface explodes that recurring cost is the thing that actually threatens the unit economics. A custom chip aimed squarely at inference is therefore not a vanity project or a research flex. It is OpenAI attacking the largest variable cost in its business at the root, trying to bend its cost-per-token curve below what it pays renting Nvidia GPUs. If Jalapeño lands anywhere near its claims, the payoff is not faster benchmarks, it is gross margin.

    The performance-per-watt claim, though, deserves the most skeptical reading in the room. OpenAI says Jalapeño will deliver performance per watt “substantially better” than current state-of-the-art, but it has not finalized the numbers, has not said which chips it tested against, on what tasks, or under what conditions, and the full technical report is somewhere in the indefinite “coming months.” These are self-reported figures from a company with an enormous interest in convincing the market it has a credible alternative to Nvidia. Hock Tan’s line that the chip is “as good as” Blackwell and Google’s TPUs is a CEO talking his own book in an interview, not a measured result. The honest posture is to treat the figures as marketing until the technical report lands. A chip running engineering samples in a lab at target frequency is real progress, but it is a very long way from a chip that holds those numbers across a production fleet under messy real-world load.

    OpenAI left the most revealing detail out of its own press release: the report, via The Decoder, that Broadcom demanded Microsoft guarantee it will buy 40 percent of the chips to secure the first phase. That single sentence tells you who is actually carrying the risk. Building gigawatt-scale custom silicon is brutally capital-intensive, and Broadcom is not willing to commit manufacturing capacity on the strength of OpenAI’s demand alone. It wants a balance sheet behind the order, and Microsoft, OpenAI’s largest backer, is the balance sheet. That detail quietly reframes the whole “OpenAI owns the stack” narrative. OpenAI may design the chip, but the deployment is underwritten by Microsoft’s purchasing commitment, which means Microsoft also gets leverage and supply security out of an OpenAI-branded part. Ownership of the design is not the same as ownership of the risk.

    The flywheel framing is genuinely interesting and probably the most defensible strategic claim OpenAI is making. OpenAI says it used its own models to accelerate parts of the chip design and optimization, compressing a normally multi-year ASIC cycle into nine months. If that is even partly true, it is a meaningful loop: the models help design the chips, the chips run the models more cheaply, the cheaper models drive more usage and revenue, and the revenue funds the next chip. That is a compounding advantage that is hard for a pure hardware vendor to replicate and hard for a pure software lab to replicate. The catch is that nine months from design to tape-out is a claim about speed, not about whether the resulting chip is actually competitive in volume. Fast tape-out and great silicon are different achievements, and the industry has seen plenty of chips that taped out quickly and underwhelmed in production.

    Strip away the “Intelligence Processor” branding and this is a playbook we have already watched run three times. Google built TPUs, Amazon built Trainium and Inferentia, Meta built MTIA, and all of them turned to Broadcom or Marvell for the design IP that is hard to replicate in-house. OpenAI is doing the same thing with the same partner, just later and louder. The diversification arc is unmistakable: OpenAI was one of the biggest Nvidia GPU buyers on earth, and in the span of a year it has signed deals for AWS Trainium, AMD accelerators, and Cerebras inference hardware, and now its own custom ASIC. Nvidia is not in trouble, demand still vastly outstrips supply, but the era where the largest AI labs were captive single-vendor customers is clearly ending. The most intriguing wildcard is OpenAI’s own line that Jalapeño is “designed with flexibility to work with all LLMs.” That is not how you describe a chip you intend to keep entirely to yourself. It hints, however faintly, at an OpenAI that could one day rent out inference infrastructure the way it now rents models, which would put it in direct competition with the very cloud providers it currently depends on.

    Key Takeaways

    • OpenAI and Broadcom unveiled Jalapeño on Wednesday, June 24, 2026, OpenAI’s first custom AI chip and its first piece of in-house silicon after years focused on models and products.
    • The chip is branded an “Intelligence Processor” and described as the first AI accelerator in a multi-generation compute platform the two companies are building together.
    • Jalapeño is purpose-built for large language model inference, the compute-intensive work of generating responses and serving answers to users, and explicitly not for training.
    • Inference is OpenAI’s recurring cost center: every ChatGPT conversation, coding request, image generation, and agent action relies on it, making it one of the highest ongoing costs in the business.
    • Broadcom President and CEO Hock Tan and President Charlie Kawwas physically delivered the first wafer to OpenAI CEO Sam Altman and President Greg Brockman.
    • OpenAI designed the chip from scratch around its understanding of LLM fundamentals, informed by its roadmap of models, kernels, serving systems, and product needs.
    • Jalapeño is described as a blank-slate design for modern LLM inference, not a general-purpose accelerator adapted from earlier AI workloads.
    • The chip is shaped by the systems OpenAI runs daily across ChatGPT, Codex, the API, and future agentic products, while also being designed to work with current and future LLMs across the industry.
    • The stated performance goal is to combine the throughput of today’s leading AI accelerators with latency closer to the fastest specialized inference systems, suiting it for interactive LLM products at scale.
    • OpenAI frames this as its full-stack advantage: it designs frontier models, builds products on top of them, and now designs the chip architecture, kernels, memory systems, networking, scheduling, and deployment systems underneath.
    • OpenAI claims Jalapeño went from initial design to manufacturing tape-out in just nine months.
    • The companies call it what they believe to be the fastest ASIC development cycle ever achieved in high-performance advanced semiconductors, against a backdrop of typically multi-year timelines.
    • OpenAI used its own AI models to accelerate parts of the chip design and optimization process, which it credits for the speed.
    • OpenAI frames the result as a flywheel: the same models served to users help improve the infrastructure that runs future models, lowering compute cost across the industry.
    • Engineering samples of Jalapeño are already running ML workloads in the lab at production target frequency and power.
    • Among the workloads running on the samples is OpenAI’s GPT-5.3-Codex-Spark model.
    • GPT-5.3-Codex-Spark currently runs on Cerebras hardware, which also specializes in inference, per The Decoder.
    • OpenAI says early testing shows Jalapeño will deliver performance per watt “substantially better” than current state-of-the-art hardware.
    • That performance-per-watt claim is self-reported and lacks independent verification; OpenAI has not said which chips it tested against, on what tasks, or under what conditions.
    • OpenAI says it is still measuring final performance and has promised a detailed technical report in the coming months.
    • The architecture reduces data movement and balances compute, memory, and networking resources to push realized utilization much closer to theoretical peak performance.
    • Jalapeño is an ASIC, which experts say is less flexible than Nvidia’s GPU but less expensive and tailorable to specific AI tasks.
    • Broadcom contributes silicon implementation and networking technologies, including its Tomahawk networking silicon, to bring the platform to large-scale production.
    • Canadian electronics manufacturer Celestica provides board, rack, and system integration expertise and will build the server systems.
    • The chips are manufactured by Taiwan’s TSMC, the world’s leading advanced semiconductor foundry, after OpenAI sent over the design.
    • Both the chips and the Celestica-built server systems will be used only by OpenAI, not sold to outside customers.
    • OpenAI plans to deploy Jalapeño at gigawatt scale by the end of 2026, with expansion in the years ahead, as the first step in a multi-generation plan.
    • Hock Tan said gigawatt-scale data center deployment will happen with Microsoft and other partners beginning in 2026.
    • The Decoder reported Microsoft is expected to buy 40 percent of the chips, with Broadcom reportedly demanding Microsoft guarantee that share to secure the first phase.
    • Broadcom CEO Hock Tan told Reuters that Jalapeño is as good as Nvidia’s Blackwell chips and the TPUs designed by Alphabet’s Google.
    • In October 2025, after 18 months of working together, OpenAI and Broadcom went public with plans to develop and deploy racks of OpenAI-designed chips starting late this year; CNBC framed the unveiling as coming eight months after that deal.
    • The prior OpenAI-Broadcom plan ultimately aimed at 10 gigawatts of custom AI accelerator capacity, with deployments expected between 2026 and 2029.
    • Estimates suggest OpenAI’s broader infrastructure plans could eventually involve around 26 gigawatts of computing capacity across custom chips, Nvidia hardware, and other accelerators.
    • OpenAI has been one of the biggest buyers of Nvidia’s GPUs since kickstarting the generative AI boom in 2022, but explosive demand has pushed it to seek other sources of advanced silicon.
    • Earlier in 2026 OpenAI struck a deal with Amazon Web Services that includes use of AWS Trainium chips, and has also signed agreements with AMD and with Cerebras, which held its IPO in May.
    • The move is widely characterized as OpenAI diversifying away from and reducing dependence on Nvidia while creating an alternative to its GPUs.
    • OpenAI’s stated goals with the chip are to reduce costs, improve energy efficiency, secure long-term computing supply, and gain more control over the infrastructure powering its services.
    • Broadcom shares climbed about 2 percent following the announcement, are up roughly 10 percent year-to-date in 2026, and have multiplied almost sevenfold since the end of 2022.
    • To build in-house chips, Meta, Amazon, and Google have turned to firms like Broadcom and Marvell for design services and IP that are hard to replicate internally; Reuters first reported OpenAI was exploring its own chip in 2023, and sources told Reuters in April 2026 that Anthropic is weighing its own AI chip.
    • Broadcom’s margin on custom AI chips is currently lower than on products like networking switches due to AI-driven high-bandwidth memory demand; Tan said SK Hynix and Samsung Electronics supply Broadcom with memory chips.

    Detailed Summary

    A blank-slate chip built only for inference

    Jalapeño is OpenAI’s first so-called Intelligence Processor, and the company is emphatic that it is not a repurposed general-purpose accelerator. It was designed from a blank slate specifically for modern large language model inference, the job of crunching data to answer a user’s query rather than the separate, bursty work of training a model. OpenAI says it designed the chip from scratch around its own deep understanding of LLM fundamentals, informed by its roadmap of models, kernels, serving systems, and product needs, drawing on the systems it runs every day across ChatGPT, Codex, the API, and future agentic products. The stated objective is to fuse the raw power and throughput of today’s leading AI accelerators with latency closer to the fastest specialized inference systems, which would make Jalapeño particularly well suited to interactive products used at scale. Notably, OpenAI also says the chip is designed with flexibility to work with all LLMs across the industry, not only its own, a claim that sits a little oddly next to its plan to keep the hardware entirely in-house.

    The full-stack flywheel and AI designing its own silicon

    OpenAI is selling Jalapeño as proof of a full-stack advantage. The argument is that because OpenAI now develops frontier models, builds products on top of them, and designs the infrastructure underneath them, including chip architecture, kernels, memory systems, networking, scheduling, deployment systems, and the product experience, every layer can be optimized around the same goal of making its models faster, more reliable, and cheaper. OpenAI describes this as a flywheel: better infrastructure drives compute efficiency, which enables better training and serving, which powers more capable models, which become better products, which drive more usage and revenue, which funds the next generation of infrastructure. The most striking piece of that loop is that OpenAI used its own AI models to accelerate parts of the chip’s design and optimization. The company’s framing is direct: if AI can help engineers design better chips faster, it can lower the cost of compute across the industry. That self-referential loop is the part of the announcement that is genuinely novel rather than a rerun of an existing hyperscaler playbook.

    Nine-month tape-out and the partner stack

    OpenAI claims it took roughly nine months to go from initial design to manufacturing tape-out, and calls this what it believes to be the fastest ASIC development cycle ever achieved in high-performance advanced semiconductors, against an industry norm measured in years. It credits deep software-hardware co-development, Broadcom’s silicon implementation expertise, and the use of its own models to compress the schedule. The work is split across a clear partner stack: OpenAI provides the architecture and AI-specific requirements, Broadcom contributes silicon implementation and networking technology, including its Tomahawk networking silicon, and Celestica handles boards, racks, and system integration, building the actual server systems. Once the design was complete, OpenAI sent it to TSMC in Taiwan, the world’s leading advanced foundry, for manufacturing. Crucially, both the chips and the systems built around them are for OpenAI’s exclusive use; they are not products being sold to outside customers.

    Performance claims that nobody can check yet

    OpenAI says early testing shows Jalapeño will deliver performance per watt substantially better than current state-of-the-art hardware, with an architecture that reduces data movement and balances compute, memory, and networking to push realized utilization much closer to theoretical peak. Hardware program lead Richard Ho said the team optimized around the kernels, memory movement, networking, and serving patterns that matter most for frontier models, and that the chip will execute key workloads close to the hardware’s theoretical limits. He told Reuters it will be performant on what he thinks will be all kinds of future LLM iterations. The important caveat is that none of this is verifiable. OpenAI is still measuring final performance, has not finalized the numbers, and has not disclosed which chips it benchmarked against, on what tasks, or under what conditions, with the technical report only promised in the coming months. As The Decoder put it bluntly, these are self-reported numbers, unverifiable for now, that should not be taken at face value. Broadcom CEO Hock Tan’s separate claim to Reuters that the chip is as good as Nvidia’s Blackwell and Google’s TPUs is similarly an unverified assertion from an interested party.

    Gigawatts, Microsoft’s 40 percent, and who carries the risk

    Jalapeño is the opening move in a much larger infrastructure buildout. Initial deployment is targeted for the end of 2026 at gigawatt scale, expanding over multiple generations. Tan said the gigawatt-scale data centers will come online with Microsoft and other partners beginning in 2026. The deal traces back to October 2025, when, after 18 months of collaboration, OpenAI and Broadcom went public with plans to deploy racks of OpenAI-designed chips, ultimately aiming for 10 gigawatts of custom accelerator capacity with deployments expected between 2026 and 2029. Broader estimates put OpenAI’s total infrastructure ambition at around 26 gigawatts across custom chips, Nvidia hardware, and other accelerators. The detail that cuts through the optimism comes from The Decoder: Microsoft is expected to buy 40 percent of the chips, and Broadcom reportedly demanded that Microsoft guarantee that purchase to secure the first phase. That guarantee shows that the financial risk of this buildout is not OpenAI’s alone; it rests heavily on its largest backer’s balance sheet.

    The Nvidia diversification arc and Broadcom’s windfall

    Jalapeño is the clearest signal yet of OpenAI loosening its dependence on Nvidia. OpenAI has been one of the biggest buyers of Nvidia GPUs since it kickstarted the generative AI boom in 2022, but demand has exploded past what any single vendor can supply. Within 2026 alone, OpenAI has struck a deal with AWS that includes Trainium chips, signed agreements with AMD and with Cerebras, which held its IPO in May, and now rolled out its own ASIC. The pattern mirrors what Meta, Amazon, and Google already did, all of them leaning on firms like Broadcom and Marvell for design IP that is hard to build in-house, and Anthropic is reportedly weighing the same move, per sources who spoke to Reuters in April 2026. Broadcom is the obvious beneficiary, with shares up about 2 percent on the news, up roughly 10 percent in 2026, and up nearly sevenfold since the end of 2022. Even so, Tan noted that the AI-driven surge in high-bandwidth memory demand makes Broadcom’s margin on custom AI chips lower than on products like networking switches, with SK Hynix and Samsung Electronics supplying the memory.

    Notable Quotes

    “The world is moving to a compute-powered economy.”

    Greg Brockman, President and Co-Founder of OpenAI, framing the launch as a broad economic shift

    “Jalapeño is part of our long-term full-stack infrastructure strategy to make compute more abundant, resulting in AI which is faster, more reliable, more affordable for people and businesses, and can be used to solve more important problems. By designing more of the stack ourselves, we can serve more intelligence with greater efficiency and keep pushing advanced AI toward broader access.”

    Greg Brockman, President and Co-Founder of OpenAI, on the full-stack rationale for building its own chip

    “Jalapeño was designed from the ground up for LLM inference using detailed insights from our close collaboration with OpenAI researchers.”

    Richard Ho, who leads OpenAI’s hardware program, describing the chip as purpose-built rather than adapted

    “We optimized the architecture around the kernels, memory movement, networking, and serving patterns that matter most for frontier AI models. Based on early testing, Jalapeño will efficiently execute our most important workloads close to the hardware’s theoretical limits.”

    Richard Ho, who leads OpenAI’s hardware program, on the architecture’s optimization targets and early performance

    “It will be performant on, we think, all kind of future iterations of LLMs.”

    Richard Ho, OpenAI hardware chief, to Reuters on the chip’s forward compatibility with future models

    “Our collaboration with OpenAI represents a fundamental commitment to scaling the physical infrastructure required for the next decade of AI.”

    Hock Tan, President and CEO, Broadcom, on the scale of the infrastructure commitment

    “This is just the beginning of a multi-generation roadmap. By co-developing our industry-leading silicon directly with OpenAI, we are enabling the deployment of gigawatt scale data centers with Microsoft and other partners beginning in 2026.”

    Hock Tan, President and CEO, Broadcom, on the multi-generation plan and 2026 gigawatt-scale deployment with Microsoft

    “The goal is to combine the power and throughput of today’s leading AI accelerators with latency closer to the fastest specialized inference systems, making Jalapeño well suited for interactive LLM products at scale.”

    OpenAI, in the press release, stating the performance objective for the chip

    “These are self-reported numbers that haven’t been finalized. Take them with a grain of salt.”

    Maximilian Schreiner, The Decoder, on the unverified performance-per-watt claim

    Jalapeño is a real chip running real workloads in a lab, but the gap between an engineering sample and a profitable production fleet is exactly where this story will be decided over the next year, and the most important numbers, the performance-per-watt figures that justify the whole effort, remain self-reported and unverified until OpenAI publishes its technical report. Read OpenAI’s full announcement here.

    Related Reading

    • OpenAI, the chip’s designer and the primary source of the announcement and quotes.
    • Broadcom, the co-developer providing silicon implementation and Tomahawk networking.
    • Celestica, which builds the boards, racks, and server systems around the Jalapeño chip.
    • ASIC (application-specific integrated circuit), what Jalapeño is, a custom chip built for one task unlike a general-purpose GPU.
    • Nvidia Blackwell, the Nvidia architecture Broadcom’s CEO claims Jalapeño matches.

  • Krishna Rao on Anthropic Going From 9 Billion to 30 Billion ARR in One Quarter and the Compute Strategy Powering Claude

    Krishna Rao, Chief Financial Officer of Anthropic, sat down with Patrick O’Shaughnessy on Invest Like the Best for one of the most detailed public looks yet at the operating engine behind Claude. He covers how Anthropic compounded from $9 billion of run rate revenue at the start of the year to north of $30 billion by the end of Q1, why he spends 30 to 40 percent of his time on compute, the playbook for buying gigawatts of AI infrastructure across Trainium, TPU, and GPU platforms, how Anthropic prices its models, why returns to frontier intelligence keep climbing, and what the Mythos release tells us about the cyber capabilities of the next generation of Claude.

    TLDW

    Anthropic is running the most compute fungible frontier lab in the world, with active deployments across AWS Trainium, Google TPU, and Nvidia GPU, and an internal orchestration layer that lets a chip serve inference in the morning and run reinforcement learning the same evening. Krishna Rao explains the cone of uncertainty that governs gigawatt scale compute procurement, the floor Anthropic refuses to drop below on model development compute, the Jevons paradox unlock from cutting Opus pricing, the 500 percent annualized net dollar retention from enterprise customers, the layer cake of long term deals with Google, Broadcom, Amazon, and the recent xAI Colossus tie up in Memphis, the phased release of the Mythos model in response to spiking cyber capabilities, the internal use of Claude Code to produce statutory financial statements and run a Monthly Financial Review skill, and why the team believes scaling laws are alive and well. The interview also covers fundraising history through Series D and Series E, the $75 billion already raised plus another $50 billion coming, talent density beating talent mass during the Meta poaching wave, and Rao’s belief that biotech and drug discovery represent the most exciting frontier for AI.

    Key Takeaways

    • Anthropic entered the year with about $9 billion of run rate revenue and ended the first quarter with north of $30 billion of run rate revenue, a more than 3x leap driven by model intelligence gains and the products built around them.
    • Compute is described as the lifeblood of the company, the canvas everything else is built on, and the most consequential class of decisions Rao makes. Buy too much and you go bankrupt. Buy too little and you cannot serve customers or stay at the frontier.
    • Rao spends 30 to 40 percent of his time on compute, even today, and the leadership team meets repeatedly on both procurement and ongoing compute allocation.
    • Anthropic is the only frontier language lab actively using all three major chip platforms in production: AWS Trainium, Google TPU, and Nvidia GPU. It is also the only major model available on all three clouds.
    • Flexibility is the central design principle. Anthropic builds flexibility into the deals themselves, into the orchestration layer that maps workloads to chips, and into compilers built from the chip level up.
    • The cone of uncertainty frames procurement. Small differences in weekly or monthly growth compound into wildly different two year outcomes, so the team plans across a range of scenarios rather than a single point estimate, and ranges toward the upper end while protecting downside.
    • Compute allocation across the company sits in three buckets: model development and research, internal employee acceleration, and external customer serving. A non negotiable floor protects model development even when customer demand is tight.
    • Anthropic estimates that if it cut off internal employee use of its own models, the freed compute could serve billions of dollars of additional revenue. It chooses not to, because internal use compounds into better future models.
    • Intelligence is multi dimensional, not a single IQ score. Anthropic measures real world capability through customer feedback, long horizon task performance, tool use, computer use, and speed at agentic tasks, not just leaderboard benchmarks that have largely saturated.
    • Each Opus generation, 4 to 4.5 to 4.6 to 4.7, delivers both capability improvements and an efficiency multiplier on token processing. New models often serve customers at a fraction of the prior cost while doing more.
    • Reinforcement learning is described as inference inside a sandbox with a reward function, so model efficiency gains directly improve internal RL throughput. The flywheel is tightly coupled.
    • Over 90 percent of code at Anthropic is now written by Claude Code, and a large share of Claude Code itself is written by Claude Code.
    • Anthropic shipped roughly 30 distinct product and feature releases in January and the pace has accelerated since.
    • Scaling laws, in Anthropic’s internal data, are alive and well. The team holds itself to a skeptical scientific standard and still does not see them slowing down.
    • Anthropic recently signed a 5 gigawatt deal with Google and Broadcom for TPUs starting in 2027, plus an Amazon Trainium agreement for up to 5 gigawatts, totaling more than $100 billion in commitments. A significant portion lands this year and next year.
    • A new partnership for capacity at the xAI Colossus facility in Memphis was announced just before the interview, aimed at expanding consumer and prosumer capacity.
    • Pricing has been remarkably stable across Haiku, Sonnet, and Opus. The biggest deliberate change was lowering Opus pricing, which produced a textbook Jevons paradox: consumption rose far faster than the price drop, and the new Opus 4.6 and 4.7 slot in at the same price point.
    • Mythos is the first model Anthropic chose to release in a phased way because of a sharp spike in cyber capability. In an open source codebase where a prior model found 22 security vulnerabilities, Mythos found roughly 250.
    • The Mythos release framework focuses on defensive use first, expands access over time, and is presented as a template for future capability spikes.
    • Anthropic now sells to 9 of the Fortune 10 and reports net dollar retention above 500 percent on an annualized basis. These are not pilots. Rao describes signing two double digit million dollar commitments during a 20 minute Uber ride to the studio.
    • The platform strategy is mostly horizontal. Anthropic will go vertical with offerings like Claude for Financial Services, Claude for Life Sciences, and Claude Security where it can demonstrate the model’s capabilities, but expects most application value to accrue to customers building on top.
    • Investors raised over $75 billion in equity since Rao joined, with another $50 billion in commitments tied to the Amazon and Google deals. Capital intensity is real, but the raises fund the upper end of the cone of uncertainty more than they fund current losses.
    • The Series E close coincided with the day the DeepSeek news broke, forcing investors to reassess their AI thesis in real time. Anthropic closed the round anyway.
    • Inside finance, Claude now produces statutory financial statements for every Anthropic legal entity, with a human checker. A library of more than 70 finance specific skills underpins workflows.
    • A custom Monthly Financial Review skill produces a 90 to 95 percent ready monthly close report, so leadership discussion shifts from reconciling numbers to debating implications.
    • An internal real time analytics platform called Anthrop Stats compresses weekly insight cycles from hours to about 30 minutes.
    • The biggest token user inside Anthropic’s finance team is the head of tax, focused on tax policy engines and workflow automation. The most senior people, not the youngest, are leading internal adoption.
    • Talent density beats talent mass. When Meta and others ran aggressive offer waves, Anthropic lost two people while peer labs lost dozens.
    • All seven Anthropic co founders remain at the company, as does most of the first 20 to 30 employees, which Rao credits to a collaborative, transparent, debate friendly culture and a real culture interview that can veto otherwise top tier candidates.
    • Dario Amodei holds an open all hands every two weeks, writes a short prepared document, and takes unscripted questions from anyone at the company.
    • AI safety investments in interpretability and alignment have a commercial side effect. Looking inside the model helps Anthropic build better models, and enterprises selling sensitive workloads want to trust the lab they hand customer data to.
    • Anthropic explicitly identifies as America first in its approach to model development, and engages closely with the US administration on capability releases such as Mythos.
    • The longer term product vision is the virtual collaborator: an agent with organizational context, access to the company’s tools, persistent memory, and the ability to work on ideas, not just tasks, over long horizons.
    • CoWork, Anthropic’s extension of the Claude Code paradigm into general knowledge work, is being adopted faster than Claude Code itself when indexed to the same point in its launch curve.
    • Anthropic’s product teams ship daily, with a fleet of agents working across the company on specific tasks. Everyone effectively becomes a manager of agents.
    • The dominant downside risks to Anthropic’s high end forecast are slower customer diffusion of model capability into real workflows, scaling laws flattening unexpectedly, and Anthropic losing its position at the frontier.
    • Rao is most excited about biotech and healthcare outcomes, especially the prospect that AI could push drug discovery and lab throughput up 10x or 100x, turning currently incurable diagnoses into treatable ones within a patient’s lifetime.

    Detailed Summary

    Compute as Lifeblood and the Cone of Uncertainty

    Rao opens with the claim that compute is the most important resource at Anthropic, and the most consequential decision class in the company. You cannot buy a gigawatt of compute next week. You have to anticipate demand a year or two in advance, and the cost of being wrong in either direction is high. Buy too much and the unit economics collapse. Buy too little and you cannot serve customers or stay at the frontier, which are described as the same failure mode. To navigate this, the team uses a cone of uncertainty rather than point estimates. Small differences in weekly growth compound into vastly different two year outcomes, and Anthropic tries to position itself toward the upper end of that cone while preserving optionality. Rao notes he has had to consciously break a lifetime of linear thinking and force himself into exponential models.

    Three Chip Platforms, One Orchestration Layer

    Anthropic uses Amazon’s Trainium, Google’s TPUs, and Nvidia’s GPUs fungibly. That was not free. Adopting TPUs at scale started around the third TPU generation, when outside observers thought it was a strange choice. Anthropic invested years into compilers and orchestration so workloads can flow across chips by generation and by job type. The team works deeply with Annapurna Labs at AWS to influence Trainium roadmaps because Anthropic stresses these chips harder than almost anyone. The result is what Rao believes is the most efficient utilization of compute across any frontier lab, with a dollar of compute going further inside Anthropic than anywhere else.

    Three Buckets and the Model Development Floor

    Compute gets allocated across model development, internal acceleration of employees, and customer serving. The conversations are collaborative rather than zero sum, but there is a hard floor on model development that the company refuses to cross even if it makes customer demand harder to serve in the short term. The thesis is simple. The returns to frontier intelligence are extremely high, especially in enterprise, so cutting model investment to chase near term revenue is a bad trade. Internal employee use is also explicitly protected. Rao notes that diverting that internal usage to external customers would unlock billions of additional revenue today, but the compounding benefit of accelerating researchers and engineers outweighs that.

    Intelligence Is Multi Dimensional

    Rao pushes back hard on the IQ framing of model progress. Benchmarks saturate quickly, and the real signal comes from how customers actually use the models. Anthropic looks at long horizon task completion, tool use, computer use, and time to result on agentic tasks. Two equally capable agents who differ only in speed produce dramatically different value, because the faster one compounds into more attempts and more outcomes. Frontier model leaps are also fuel efficient. The sedan to sports car analogy breaks down because each Opus generation, 4 to 4.5 to 4.6 to 4.7, delivers a step up in capability and a multiplier on per token efficiency.

    From 9 Billion to 30 Billion ARR in One Quarter

    The headline number for the quarter is a leap from about $9 billion of run rate revenue to over $30 billion, accomplished without onboarding a corresponding step up in compute, because new compute lands on ramps locked in 12 months prior. Rao attributes the leap to model capability gains, products that surface that intelligence in usable form factors, and an enterprise customer base that pulls more workloads onto Claude as each generation unlocks new use cases. Coding started the wave with Sonnet 3.5 and 3.6, and the same pattern is now playing out elsewhere in the economy.

    Recursive Self Improvement and Talent Density

    Over 90 percent of Anthropic’s code is now written by Claude Code, including most of Claude Code itself. Rao describes this as a structural reason to keep allocating internal compute to employees even when external demand is hungry. Recursive self improvement is not happening through models that need no humans. It is happening through researchers who set direction and use frontier models to compress months of work into days. Talent density beats talent mass. When Meta and other labs went after Anthropic researchers with very large packages, Anthropic lost two people while peer labs lost dozens.

    Procurement Strategy and the Layer Cake

    Compute lands as a layer cake. Last month Anthropic signed a 5 gigawatt TPU deal with Google and Broadcom starting in 2027, alongside an Amazon Trainium agreement for up to 5 gigawatts. The total is north of $100 billion in commitments. A new tie up with xAI’s Colossus facility in Memphis was announced just before the interview, intended for nearer term capacity to support consumer and prosumer growth. Anthropic evaluates near term and long term compute deals against the same set of variables: price, duration, location, chip type, and how efficiently the team can run it. The relationships are deeper than procurement. The hyperscalers are also distribution channels for the model.

    Platform First, Selective Vertical Bets

    Rao describes Anthropic as a platform first business, with most expected value accruing to customers building on the platform. The team will only go vertical when it can either demonstrate capabilities that are skating to where the puck is going, like Claude Code did before the models could fully support it, or when it wants to set a template for an industry vertical, as with Claude for Financial Services, Claude for Life Sciences, and Claude Security. He acknowledges that surprise capability jumps make customers anxious about the platform competing with them, and frames Anthropic’s mitigation as deeper partnerships, early access programs, and an emphasis on accelerating customer building rather than disintermediating it.

    Pricing, Jevons Paradox, and Return on Compute

    Pricing across Haiku, Sonnet, and Opus has been stable. The notable exception is Opus, which Anthropic deliberately repriced lower when launching Opus 4.5 because Opus class problems were being squeezed into Sonnet workloads. Efficiency gains made it possible to serve Opus profitably at the new level. The consumption response was a classic Jevons paradox, with usage rising far more than the price reduction would have predicted, and Opus 4.6 then slotted in at the same price with a capability bump. Margins are not framed as a per token markup. Compute is fungible across model development, internal acceleration, and customer serving, so Anthropic measures return on the entire compute envelope rather than software style variable cost per call.

    Fundraising, DeepSeek, and Capital Intensity

    Rao joined while Anthropic was closing its Series D, mid frontier model launch and during the FTX share liquidation. Investors initially questioned whether Anthropic needed a frontier model, whether AI safety and a real business could coexist, and why the sales team was so small. The Series E closed the same day the DeepSeek news broke, with markets violently re pricing AI in real time. Since Rao joined, Anthropic has raised over $75 billion, with another $50 billion tied to the Amazon and Google compute deals. The reason for the size of the raises is the cone of uncertainty, not current losses. Returns on compute today are described as robust.

    Mythos, Cyber Capability, and Phased Releases

    The Mythos release marks the first time Anthropic shipped a model under a deliberately phased rollout because of a specific capability spike. Cyber is the dimension that spiked. Where a prior model found 22 vulnerabilities in an open source codebase, Mythos found roughly 250. The defensive applications, automatically patching massive codebases, are genuinely valuable, but the offensive risk is real enough that Anthropic chose to release to a smaller group first and expand access over time. Rao positions this as a template for future capability spikes, not a permanent restriction. He also describes the relationship with the US administration as cooperative, including the Department of War interaction, with Anthropic supporting a regulatory framework that does not strangle innovation but takes responsibility seriously.

    Claude Inside Finance

    Anthropic’s finance team is one of the strongest internal case studies. Statutory financial statements for every legal entity are produced by Claude, with a human reviewer. A skill library of more than 70 finance specific skills underpins a Monthly Financial Review skill that drafts the monthly close at 90 to 95 percent ready, so leadership meetings shift from explaining the numbers to discussing what to do about them. An internal analytics platform called Anthrop Stats compresses weekly insight cycles from hours to 30 minutes. The biggest internal token user in finance is the head of tax, building policy engines, which Rao highlights as evidence that adoption is driven by the most senior people, not just younger engineers.

    Culture, Co Founders, and the Race to the Top

    Seven co founders should not, on paper, work as a leadership group. Rao argues it works because the culture was set early around collaboration, intellectual honesty, transparency, and humility. The culture interview is a real veto, not a checkbox. Dario Amodei runs an all hands every two weeks with a short written piece followed by unscripted questions, and decisions, once made, get clean alignment rather than residual politics. Anthropic frames its approach as a race to the top, where being a model for how to build the technology responsibly is itself a recruiting and retention advantage.

    The Virtual Collaborator and the Frontier Ahead

    The product vision Rao describes is the virtual collaborator. Not just a smarter chatbot, but an agent with organizational context, access to the company’s tools, memory, and the ability to work on ideas over long horizons. Coding was the first domain to feel this, but CoWork, Anthropic’s extension of the Claude Code pattern into general knowledge work, is being adopted faster than Claude Code was at the same age. Product development inside Anthropic already looks different. Teams ship daily, with fleets of agents working across the company, and individual humans increasingly act as managers of those fleets.

    Downside Risks and What Excites Him Most

    The three risks Rao names if asked to do a premortem on a softer year are slower customer diffusion of model capability into real workflows, scaling laws unexpectedly flattening, and Anthropic losing its frontier position to competitors. None of these are observed today, but he is unwilling to claim them with certainty. On the upside, he is most excited about biotech and healthcare. Lab throughput rising 10x or 100x, paired with AI assisted clinical workflows, could turn currently incurable diagnoses into treatable ones within a patient’s lifetime. That is the outcome he wants the technology to chase.

    Thoughts

    The most consequential structural point in this interview is the framing of compute as a single fungible resource pool measured by return on the entire envelope, not as a variable cost per inference call. That accounting shift, if you accept it, breaks most of the bear cases about AI lab unit economics. The bear argument almost always assumes that a token served to a customer is the only thing the chip did that day. Rao’s version is that the same fleet trains models in the morning, runs reinforcement learning at lunch, serves customers in the afternoon, and accelerates internal engineers in the evening. If even half of that is real, the right comparison is total compute spend versus total enterprise value created by the platform, and on that ratio Anthropic looks structurally strong rather than weak.

    The Jevons paradox on Opus pricing is the most actionable insight for anyone running an AI product. Most teams default to either chasing premium pricing on the newest model or undercutting to chase volume. Anthropic did something more disciplined: it left Sonnet and Haiku alone, dropped Opus when efficiency gains made it serveable, and watched aggregate usage rise faster than the price cut. The lesson is that frontier model pricing is not really a price problem. It is a capability access problem, and elasticity around the right tier is much higher than the standard SaaS playbook implies.

    The Mythos cyber jump deserves more attention than it has gotten. Going from 22 to 250 vulnerabilities found in the same codebase is the kind of capability discontinuity that genuinely changes the regulatory calculus. Anthropic is signaling that it can identify these discontinuities ahead of release and choose a deployment shape that respects them. Whether peer labs adopt similar discipline is the open question. Anthropic’s race to the top framing assumes they will be forced to. The competitive market may say otherwise.

    The hiring data point is the most underrated investor signal. Two departures while peer labs lost dozens, during the most aggressive talent war in tech history, is not a culture poster. It is a structural advantage that compounds every time another lab tries to buy its way to the frontier. Money can be matched. Conviction in the mission, transparent leadership, and a culture interview that can veto otherwise stellar candidates cannot. If you believe scaling laws hold, talent retention at this density is one of the few moats that actually scales with capital.

    Finally, the most interesting personal admission is that Krishna Rao, a finance leader trained at Blackstone and Cedar, is openly telling investors that linear thinking is the failure mode he had to break out of. The companies that pattern match this moment to prior technology waves are mispricing it, in both directions. The cone of uncertainty Anthropic uses internally is the right metaphor for everyone else too. If you are forecasting AI as if it is cloud in 2010, you are almost certainly wrong, and the magnitude of the error is much larger than it would be in any prior era.

    Watch the full conversation with Krishna Rao on Invest Like the Best here.