ICE plans compute futures contracts as Wall Street races to turn GPU power into a tradable commodity

At a glance:

• ICE and Ornn to launch cash-settled futures tied to GPU compute costs.
• CME Group announced similar contracts days earlier, partnering with Silicon Data.
• GPU rental prices have surged, highlighting need for risk management tools.

ICE and Ornn's partnership

Intercontinental Exchange, the owner of the New York Stock Exchange, is partnering with Ornn, a financial-infrastructure firm, to develop cash-settled futures contracts based on GPU computing costs. The move, announced on Monday, pairs a major exchange operator with a startup that has built real-time pricing indexes for compute. Ornn's flagship product, the Ornn Compute Price Index, tracks live spot prices for GPUs including Nvidia's H100, H200, and B200 chips, and is available on the Bloomberg Terminal. The index draws on transaction data from live GPU markets and has attracted over 400 data centre operators, investors, and AI companies.

Trabue Bland, senior vice president of futures markets at ICE, emphasized the market's need for standardized pricing. "The compute market," Bland said, "is in desperate need of a globally accepted pricing mechanism and risk management tool" as AI becomes a central economic driver. The futures contracts will be US dollar-denominated and cash-settled, providing a way for AI firms and cloud providers to hedge against volatile costs that have accompanied Big Tech's $650 billion capital expenditure surge projected for 2026.

The competitive landscape with CME

ICE is not alone in this venture. CME Group, the world's largest derivatives exchange, announced its own compute futures on 12 May, partnering with Silicon Data to create products based on the Silicon Data H100 Rental Index. This index tracks daily rental rates for high-end GPUs used in AI training. The simultaneous moves by ICE and CME signal that institutional belief in compute-as-commodity has reached a tipping point, reminiscent of the early days of energy futures when exchanges raced to establish benchmarks for oil and natural gas.

The competition extends beyond these two giants. Architect Financial Technologies launched exchange-traded perpetual futures on GPU and RAM prices through its AX platform in January, and Kalshi offers contracts for wagering on Nvidia GPU compute prices. However, ICE and CME bring deep institutional liquidity, regulatory credibility, and clearing infrastructure that large-scale GPU-as-a-service providers and their customers will demand, potentially setting the reference price for the industry.

Why the AI economy needs compute futures

Kush Bavaria, co-founder and CEO of Ornn, stated bluntly that compute "has grown into a trillion-dollar market, yet it still lacks the pricing and risk-transfer infrastructure that every other major commodity relies on." This gap has led to significant volatility; for instance, Ornn's index showed Nvidia's Blackwell spot rental price surging 48% from $2.75 to $4.08 per GPU-hour between mid-February and mid-April 2026. For AI companies running multi-million-dollar training jobs, such swings can derail budgets, while cloud providers and data centre operators face similar exposure.

A functioning futures market would allow participants to lock in forward prices, transfer risk, and plan capital expenditure with greater certainty. It would also generate transparent price signals currently absent, giving investors, analysts, and policymakers clearer insights into cost trends. This is crucial as AI transitions from research labs to core economic infrastructure, with surging semiconductor demand already reshaping supply chains and driving record investment.

Broader implications for financial markets

The emergence of compute futures reflects a deeper financialization of AI inputs, akin to how energy, metals, and agricultural products were commoditized in prior decades. As AI becomes ubiquitous, the resources powering it are being transformed into standardized financial assets. This creates benchmarks that can underpin lending decisions, insurance products, and investment strategies tied to AI infrastructure. For example, a bank financing a new data centre could use compute futures to assess projected revenue against forward GPU prices, similar to how energy lenders use oil futures for drilling projects.

Ornn has designed its futures with Asian-style settlement to align with the real-time consumption of compute, settling on the arithmetic average of daily index values across the contract's tenor rather than a single expiry-day price. This structure addresses the challenge that compute is a "flow commodity" that cannot be stored. Whether ICE or CME captures the market will depend on liquidity, GPU type coverage, and institutional trust in index providers, but the trajectory is clear: computing power is being standardized and traded as a financial asset.

Challenges and the road ahead

Despite the promise, compute futures face hurdles. The nature of compute as a flow commodity requires innovative contract designs, which Ornn has tackled with its Asian-style settlement. Success will hinge on attracting sufficient liquidity and establishing trusted benchmarks. The exchange that sets the reference price early on, much like ICE Brent and CME WTI did for oil, could dominate the market.

For an industry accustomed to opaque, bilateral GPU procurement deals, this shift to a transparent, exchange-traded market is significant. It promises to reduce costs, mitigate risks, and provide much-needed price discovery for the AI economy. As Wall Street races to turn GPU power into a tradable commodity, the foundations of a new financial market are being laid, with far-reaching implications for technology, finance, and the broader economy.