The U.S. Commodity Futures Trading Commission issued a request for comment on August 19 to establish rules for trading compute derivatives contracts, a watershed moment for formalizing GPU capacity as a financial asset. CFTC Chairman Michael Selig stated: 'America cannot win the AI race without a robust derivatives market for compute.' The comment period runs 60 days from Federal Register publication, with responses due by October 20. The agency seeks feedback on cash-market size and liquidity, manipulation risks, customer protection, and the feasibility of perpetual compute futures.
Exchanges have moved ahead of the regulator. CME Group and Silicon Data announced plans to list two contracts on October 5, pending regulatory approval. The contracts will track hourly GPU rental costs for NVIDIA H100 and next-generation B200 chips—the two most critical accelerators in the current AI stack. This represents the first standardized, cleared derivatives vehicle for compute pricing, filling a gap that has left GPU capacity priced via opaque spot and forward markets.
For infrastructure architects and financiers, this institutionalizes compute as a commodity class alongside oil, metals, and energy. Once live, the contracts enable hedging of compute exposure, financing of datacenter buildout, and price discovery across cloud providers—critical for cost planning in training and inference workloads. The fast timeline (October listing, 60-day comment window ending October 20) suggests the CFTC intends minimal friction between public feedback and launch, underscoring how central compute derivatives are viewed in U.S. AI competitiveness policy.