Justin Halford
Founder of Voyadoc
Nvidia's 70–75% margins will erode because most of its revenue comes from companies with every incentive to cut it out. Anthropic has committed $200 billion to Google's TPU alternative, and OpenAI won 30% off its Nvidia order by threatening to do the same. Google is likely to sell TPUs externally by 2027, making Nvidia the biggest victim of AI's near-term valuation froth. x.com
Broadcom's integrated optics, electronics and packaging could lower custom-chip system costs because separate suppliers defend their margins. At similar valuations, it is preferred to Nvidia and has more room to beat expectations. Google TPU v8 shipments are estimated at 4.0M in 2026 and 9.1M in 2027, assuming Broadcom supplies the remainder after MediaTek's allocation. Anthropic financing adds value only if chip profit exceeds likely support payments and capacity costs. x.com x.com
NVIDIA faces rival chips from every Mag 7 customer, which has an incentive to cut GPU spending and improve margins on its AI buildout. If they can make a GPU 85% as good for 50% less, it is worth using instead of NVIDIA's; the shift will take time. NVIDIA makes a GPU for $5,000 and sells it to those customers for $50,000, leaving them an incentive to save 90% by building their own. x.com x.com x.com x.com
NVIDIA's liquid-cooling work with Vertiv can deliver more compute from a fixed power budget. Liquid cooling yields roughly 22% more compute and more tokens from the same power on average, a benefit modeled in an April report. NVIDIA's scale, balance sheet and command of its supply chain make it the economic winner of the AI compute cycle, with Broadcom a close-ish second. x.com x.com
Nvidia's funding of future customers in Germany further blurs how much chip demand it paid for itself. Jensen Huang said Nvidia invested almost $1B there in the last year and is investing in four more companies this week; German taxpayers are helping too. Finding buyers beyond Big Tech is smart, but the more sales come back from Nvidia's own checks, the harder genuine demand is to gauge. x.com
Nvidia's AI profits are real, but its compute demand depends heavily on fragile venture funding. Per Chamath, 1% of firms account for 90% of AI spending, and 72 cents of each dollar comes from unprofitable VC-backed firms. Venture money passes through startups to Nvidia and cloud suppliers; if the next raises fail, concentrated orders could contract together. x.com x.com x.com
Nvidia retains pricing power if its full system keeps reducing customers' cost per effective AI task. Strong AI demand alone does not ensure profits across the chain: rising chip, memory and optical costs burden buyers, while tighter financing and slow collections threaten operators. Large customers developing their own chips could pressure Nvidia's margins even as the market grows. x.com
Falling AI token and compute prices are a perfect storm for Nvidia. Token costs are down about 30% since the start of June, falling on 15 of the past 18 days, while compute rental prices are 16% off their recent highs. The bull argument that this doesn't matter for Nvidia because the Silicon Token LLM index only tracks spot prices does not hold. x.com
Nvidia faces growing doubt over how much it must finance its own customers to sustain AI demand. It fell 5.1%, erasing about $256 billion, on reports it could backstop up to $250B for an OpenAI data-center project, effectively funding OpenAI to buy its products. CXMT's surging debut added worry about Chinese chip competition. Apple overtook Nvidia as the most valuable company as markets separate AI beneficiaries from firms taking aggressive financial risks to keep the cycle going. x.com
AI infrastructure bottlenecks could give eight stocks as much as 18x upside. Nvidia supplies frontier-model GPUs; Broadcom and Marvell supply custom silicon and networking; TSMC makes advanced chips; Micron supplies sold-out HBM; Vertiv provides power and cooling; Palantir monetizes enterprise AI; Alphabet spans TPUs, research, cloud and apps. The list tracks where AI infrastructure spending is flowing. x.com x.com
Nvidia and other accelerator makers now face a finite semiconductor profit pool and must fight their supply chains for incremental gross margin, pushing down dollars per GW after years of inflation. Economic surplus needs to move from silicon to users; otherwise the inflation from semis and backlash against data centers will lead rate markets and policy to end this trade. Expecting $150-175B of AI lab ARR to justify $1T+ of capex sounds insane. x.com x.com