Financial and technology firms are racing to establish a futures market for AI compute capacity to bring transparency and risk management to the sector. CME Group plans to launch two compute futures contracts on October 5, 2026, representing monthly rentals of Nvidia H100 and Blackwell B200 chips. Meanwhile, startup OneChronos is developing a marketplace using combinatorial auctions designed with Nobel laureate Paul Milgrom to address challenges like compute's lack of fungibility and storability.
AI compute futures market launch
- ▪Each CME Group compute futures contract will represent a month's worth of rent for either the Nvidia H100 or the Nvidia Blackwell B200 graphics processing units.
- ▪CME Group announced on August 11, 2026, that it plans to launch two AI compute futures contracts on October 5, 2026, pending regulatory review.
- ▪New York-based startup OneChronos is developing a compute futures marketplace and expects it to be operational in 2026, pending approval from federal regulators.
- ▪CME Group is collaborating with Silicon Data, a company that publishes indexes tracking compute pricing, to develop its planned AI compute futures contracts.
Compute as investable asset class
- ▪Nvidia CEO Jensen Huang characterized AI compute as an investable asset class in an August 10, 2026, post regarding a $500 billion financing deal with six financial firms.
- ▪BlackRock CEO Larry Fink stated at the Milken investor conference in May 2026 that buying futures of compute would become a new asset class due to a shortage of compute power.
Commodity classification challenges
- ▪OneChronos CEO Kelly Littlepage stated that AI compute fails standard commodity measures because it is not fungible, storable, or transportable.
- ▪Pete Keavey, global head of energy and environmental products at CME Group, stated that the exchange operator aims to turn AI compute into a standardized, tradable commodity.
Combinatorial auction solution
- ▪OneChronos is working with Nobel Prize-winning economist Paul Milgrom, who previously developed wireless spectrum auctions, to design its combinatorial auction system for AI compute.
- ▪OneChronos is utilizing combinatorial auctions in its futures market to address the non-fungible, non-storable, and non-transportable characteristics of AI compute.
Market transparency benefits
- ▪Proponents expect an AI compute futures market to bring greater price transparency, new risk management tools for volatility, and increased investment into the sector.
- ▪Traditional futures markets allow companies to hedge against price swings, such as airlines purchasing oil futures or farmers selling soybean futures to manage costs.
Investment capital unlocking potential
- ▪Historically, establishing a transparent market for a commodity has unlocked investment from traders and investors who do not physically own the underlying asset.
- ▪Financial and technology firms are actively competing to transition AI compute from a metaphorical currency of the artificial intelligence age into a standardized financial asset.
Story comments
Loading comments…