ICE is working on GPU compute futures in collaboration with its partner, Ornn index.

      TL;DRICE, the proprietor of the New York Stock Exchange, is collaborating with index provider Ornn to introduce cash-settled futures contracts linked to GPU computing costs. This announcement follows a similar move by competitor CME Group, highlighting Wall Street's rush to standardize AI computing power as a tradable commodity.

      Intercontinental Exchange, the parent company of the New York Stock Exchange, is set to launch futures contracts based on computing power costs, indicating Wall Street's belief that AI infrastructure represents the next major commodity market. ICE revealed on Monday that it will join forces with Ornn, a financial infrastructure firm that tracks GPU computing costs in real time, to create these new contracts. The futures will be denominated in US dollars, settled in cash, and will reference Ornn's indexes covering various major GPU types, pending regulatory approval.

      The collaboration pairs one of the largest exchange operators with a startup that has developed a system for compute price discovery. Ornn, officially Ornn AI Inc, provides the Ornn Compute Price Index, which monitors real-time spot prices for GPU compute across different hardware, including Nvidia's H100, H200, and B200 chips. The index, now accessible on the Bloomberg Terminal, relies on actual transaction data from live GPU markets and has garnered interest from over 400 data center operators, investors, and AI companies.

      Trabue Bland, senior vice president of futures markets at ICE, described this initiative as a response to a computing market that has outgrown its existing informal pricing methods. He pointed out that the compute market urgently requires a globally recognized pricing mechanism and risk management tool as AI transitions from research settings to a significant component of the global economy.

      The new contracts will be cash-settled, similar to energy and financial futures. This structure could provide AI companies managing extensive model training and cloud providers securing capacity a means to hedge against the fluctuating compute costs that have accompanied Big Tech's $650 billion capital expenditure growth in 2026.

      ICE is not the only entity recognizing this opportunity. CME Group, the largest derivatives exchange worldwide, launched its compute futures contracts on May 12, partnering with Silicon Data to create products based on daily GPU benchmark rental rates. CME's contracts will reference the Silicon Data H100 Rental Index, which tracks the expenses of renting high-performance GPUs for AI training.

      The simultaneous moves by these two prominent futures exchanges signal a significant institutional belief in compute as a commodity. This scenario mirrors the early days of energy futures in the 1980s when competing exchanges rushed to establish benchmark contracts for crude oil and natural gas. The exchange that secures the most liquidity early will likely dictate the reference price for the sector, akin to how ICE Brent and CME WTI determined oil prices.

      This competitive environment extends beyond the two major exchanges. Architect Financial Technologies, which partnered with Ornn in January, aims to launch exchange-traded perpetual futures on GPU and RAM prices via its AX platform. Additionally, prediction market Kalshi offers contracts that allow users to bet on Nvidia GPU compute prices. However, ICE and CME possess advantages over newer entrants, including substantial institutional liquidity, regulatory credibility, and the clearing infrastructure crucial for large-scale GPU-as-a-service providers and their clients.

      Kush Bavaria, co-founder and CEO of Ornn, emphasized the scale of the issue, stating that compute has evolved into a trillion-dollar market but still lacks the necessary pricing and risk-transfer mechanisms found in other major commodities. This deficiency leads to tangible repercussions, as GPU rental prices have shown extreme volatility. Ornn's index revealed that the Nvidia Blackwell spot rental price rose 48% from mid-February to mid-April 2026, escalating from $2.75 to $4.08 per GPU-hour. For AI companies whose model training expenses can reach tens of millions of dollars, such price fluctuations can severely impact budgets. Similar risks confront cloud providers, data center operators, and lenders funding billions in AI infrastructure projects.

      A functioning futures market would permit these stakeholders to secure forward prices, transfer risk to willing counterparties, and plan capital expenditures with greater precision. Additionally, it would generate clear price signals currently lacking in the broader market, helping investors, analysts, and policymakers better understand future computing costs.

      The advent of compute futures signifies a more profound structural change. As AI transitions from a novel technology to a fundamental economic driver, the resources that fuel it are increasingly being financialized, similar to the processes applied to energy, metals, and agricultural products in previous eras. The rising demand for advanced semiconductors has already reshaped chip supply chains and stimulated unprecedented capital investments across the tech sector.

      Futures contracts introduce a new dimension to this ecosystem, establishing standardized benchmarks that can inform lending decisions, insurance products, and investment strategies related to AI infrastructure. For instance, a bank financing a new data center could leverage compute futures to evaluate its anticipated revenue in relation to future GPU prices, much like energy lenders assess drilling projects using oil futures.

      There are challenges ahead, however. Unlike oil