Nvidia unveiled its latest Rubin GPU to CNBC on February 13, 2026 at its headquarters in Santa Clara, California.
Mark Ganley | CNBC
Computing power is emerging as a new tradable asset class, with CME Group set to launch the first futures contract tied to the operating costs of the chips that power artificial intelligence.
The exchange, in partnership with Silicon Data, plans to introduce two computing futures contracts on October 5, pending regulatory approval, giving companies and investors a way to trade and hedge the price of AI computing power, similar to oil, electricity and other commodities.
“For years, two companies buying the exact same GPU capacity could pay vastly different prices, with no way of knowing who got the better deal. Now, both companies will have a benchmark to check that with,” Silicon Data CEO Carmen Li said in a statement. “Computing futures provide the market with something never seen before: a publicly tradable reference price for the resources on which any AI system runs.”
The deal will allow buyers and sellers to trade on rental costs for Nvidia’s H100 and newer Blackwell B200 graphics processing units, based on the Silicon Data Index, which tracks hourly GPU rental prices. Each contract represents one month’s rental of an Nvidia H100.
The announcement comes as Wall Street looks for new ways to fund and gain exposure to building large-scale AI infrastructure. Nvidia is working with some of the world’s largest asset managers on an initiative that could potentially pump as much as $500 billion into AI infrastructure.
Computing futures will add a new layer to that emerging financial ecosystem. Rather than investing directly in data centers, chips, or the companies that build them, investors can gain exposure to the price of the underlying computing power itself, and AI developers and data center operators can use contracts to hedge costs and revenues.
