Startup bets that investors want to trade compute like a commodity
Get the Finance newsletter
Daily finance — markets, central banks, M&A, the prints that move money. Free.
- Ornn, backed by Andreessen Horowitz, raised a $33 million seed round to build a marketplace for trading AI compute, modeled on oil commodity markets with futures contracts that lock in prices.
- Goldman Sachs estimates $7.6 trillion will be invested globally in compute, power, and data centers between 2026 and 2031, noting the financial infrastructure to sustain that level of spending "has not yet been built."
- Nvidia's chip roadmap is a complicating factor — each new generation depreciates older GPU value, and unused compute cannot be stored like oil barrels, making standardized pricing and contracts harder to design.
- Ornn already operates under a de minimis exemption and has integrated with Bloomberg Terminal for GPU price discovery; CME plans compute futures tied to Silicon Data's benchmark and Intercontinental Exchange plans GPU futures tied to Ornn's pricing index, all pending regulatory approval.
- Ornn CEO Kush Bavaria told Axios the company does not work with Chinese AI labs, framing the venture as part of America's potential advantage over China in compute financing.
- Ornn CTO Wayne Nelms told Axios: "We want to make financing AI way more seamless."
Why it matters: If standardized compute contracts work, AI labs and GPU owners gain a hedge against Nvidia's chip depreciation cycle — but each new generation makes older benchmarks obsolete and unused GPU capacity vanishes overnight. Goldman Sachs says $7.6 trillion in compute spending by 2031 needs financial scaffolding that does not yet exist.
