Kyle Reidhead
CEO & Analyst of Milk Road/Milk Road AI
Physical AI is a compute driver the market still is not pricing, because robots need roughly 10x more compute than digital AI and must be trained before they ship. Figure committed $3.5B, expected to exceed $6B, for up to 100,000 Nvidia Vera Rubin GPUs through neocloud Nscale, with robots still in pilots. That lights up Nvidia, neoclouds such as CoreWeave and Nebius, and memory, about 29 petabytes of HBM4. Add Tesla's Optimus and others, and the buildout runs for years. x.com x.comÂ
Physical AI could be an order of magnitude bigger than digital AI, and Tesla, Nvidia, Micron and Infineon cover its hardware stack. Nvidia expects physical AI revenue to grow from $10B to $100B in a decade, and its VP of Robotics notes almost none of the world's physical work has been touched by AI. Nvidia supplies the brains, Tesla leads in robotaxis and humanoids, every robot needs Micron's memory, and Infineon's power chips are quiet winners. The trend is just starting. x.comÂ
Physical AI could open a new demand curve for Nvidia before humanoids ship in volume, as the biggest chip opportunity may first sit in data centers training robots. Figure reportedly plans about 100,000 Vera Rubin GPUs from 2027, roughly 230MW mostly for training. Per Goldman, a humanoid holds only a few thousand dollars of chips, but each robotics company chasing frontier capability may become a major compute customer. Some demand is circular; training bills pay off only if fleets grow large. … Show more
Enterprise and neocloud demand is creating a second wave of AI infrastructure spending, with tailwinds for Nvidia and Super Micro. Dell forecasts 200 GW of AI data center growth through 2030, half driven by these customers, and estimates a $1 trillion AI infrastructure opportunity over the next four years. AI servers could enter a new growth phase in 2027, with higher shipment volumes and more revenue per deployment. x.comÂ
Nvidia's neocloud orders increasingly require its own credit support. Per The Information, it is in talks to guarantee billions of Nscale debt for a 460MW data center; Nscale needs an IPO to fund $24B in Nvidia systems committed for 2026-27. Separately, a quoted article says US officials identified Alice Huang, scrutinized for possibly routing Nvidia chips to China, as a Chinese intelligence asset with direct access to Jensen Huang. This is just the tip of a very big iceberg. x.com x.com x.comÂ
Nvidia can capture more AI infrastructure spending by supplying Vera CPUs and networking alongside GPUs. Oracle outlined an initial supply of 150,000 Vera CPUs, and OpenAI says it is putting them into production for agent tool execution; reducing CPU waits could improve cost per task and GPU-memory use. Benefits from independently scaling CPUs and GPUs still need validation in utilization and completion times. x.comÂ
Lenders committing billions to neoclouds weaken the case that expansion will be starved of affordable capital. IREN secured $363M of GPU-backed financing and Nebius a $3.5B credit facility. Longer-lived, redeployable Nvidia GPUs are becoming more acceptable collateral; customer prepayments and cash flow from new contracts offer other funding sources as compute demand outpaces supply. x.comÂ
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.comÂ
AI agents' productivity over human labor should lift GPU demand and rates for neoclouds. The bottom looks in for CoreWeave, Nebius, Boost Run, Sharon AI and IREN. Gavin Baker says token prices can fall as revenue per GPU rises; paraphrasing Elon, even $38 per GPU-hour can pay off. CoreWeave's five-year fixed contracts may lag current rates; Boost Run has less capacity locked in. If SpaceX takes more growth, slower cash burn could rerate neoclouds, though smaller customers may pay less. x.com … Show more
A rising value of compute per watt should benefit AI cloud, chip, connectivity and power suppliers. Elon Musk says that value will rise over time. Nebius has Microsoft and Meta deals; Marvell benefits from custom silicon demand; Credo connects GPUs; Vistra has long-term nuclear supply deals with Amazon and Meta; Nokia's Nvidia partnership could help it capture data-center networking growth. x.com x.comÂ
Nvidia could become the weakest link in AI's circular financing economy as it takes on more customer risk. It is turning to insurers to cover customers' GPU debt, and if insurers require Nvidia to keep backstopping deals, its liabilities will keep growing on top of $245B of supply chain commitments and prepayments as of last quarter. Such large counterparty conflicts are unwise. Amazon is also looking to move $8B of Nvidia Blackwell GPUs into an SPV to take on more debt for its AI buildout. … Show more
Anthropic's customer spending and compute buildout show AI monetizing at scale, a bullish read-through for Nvidia, Broadcom, Alphabet, Amazon, Microsoft, Nebius, CoreWeave and Hut 8. Its IPO filing reports H1 revenue up 14-fold to $16.3B, 11,900+ customers spending $100K+ annually and 1,900+ spending $1M+, alongside $9B in H1 infrastructure spending and about $430B in future compute commitments. x.comÂ
