CME Group will launch two futures contracts tied to artificial intelligence computing power on October 5, pending regulatory approval, the exchange announced August 11. The contracts, developed in partnership with Silicon Data, will allow investors to trade exposure to GPU and CPU capacity as a standalone asset class.
The move marks the first time a major U.S. futures exchange has listed derivatives on compute infrastructure rather than on the chips themselves or the companies that manufacture them. CME's existing contracts track semiconductor stocks and the Nasdaq-100; these new instruments isolate the rental value of processing power. The mechanism mirrors how energy futures let traders hedge power supply independent of oil or gas prices.
Silicon Data operates a platform that aggregates and prices compute capacity across cloud providers and data centers. The company's role is to supply the underlying pricing methodology and data feed that will determine contract settlement values. CME said the contracts are designed to let data center operators, cloud providers, and AI companies hedge their exposure to compute costs, which have become volatile as demand for training infrastructure fluctuates with model development cycles.
AI infrastructure spending has accelerated sharply since late 2022. Capital expenditure on data centers and GPUs reached approximately $60 billion globally in 2025 according to industry estimates cited in prior announcements. No comparable futures market currently exists for compute as a standalone commodity; financial traders have used proxy hedges like Nvidia stock or broad semiconductor indices.
The regulatory pathway for these contracts remains open. CME submitted the proposal to the Commodity Futures Trading Commission, which has 45 days to issue initial guidance under standard procedures. The October 5 launch date is contingent on CFTC clearance. CME has listed 13 new derivative contracts in 2026 to date, according to its regulatory filings, spanning digital assets, energy, and metals.
Two compute futures contracts represent less than one percent of CME's total listed products, but the exchange is testing whether financial markets can price the marginal cost of AI infrastructure the way they price oil barrels or copper tonnes. If the contracts attract sustained trading volume, other exchanges may follow; if liquidity remains thin, CME may adjust specifications or retire the products within two years as it has done with failed launches in prior cycles.
The regulatory outcome and early trading volume will show whether compute pricing is mature enough for institutional hedging, or whether the market for it remains too fragmented and fast-moving to support standardized derivatives.
