Two new futures contracts are poised to begin trading October 5 on the NYMEX market. They “will bring much-needed hedging and investment vehicles to businesses looking to manage compute, the processing power and hardware infrastructure that machines need to train and run AI models.”
So said NYMEX parent CME Group when announcing the launch with its partner Silicon Data, whose indexes on Nvidia GPU chip rental costs underlie the contracts.
Since those organizations went public with their plans in May, “compute” has increasingly come to represent an asset class encompassing the rampant demand for processing power and infrastructure, from GPUs (graphics processing units) to data centers, amounting to trillions of dollars of capital expenditures.
“As the backbone of the digital economy, compute is the new oil of the 21st century,” said CME Group Chairman and CEO Terry Duffy, who is not alone in invoking that metaphor. Nor are CME and Silicon Data alone on the compute-product journey.
ICE SVP Trabue Bland
Also in May, Intercontinental Exchange announced with Ornn – which like Silicon Data operates a GPU marketplace – plans for GPU compute futures based on Ornn’s Compute Price Index.
“As AI has rapidly moved from research labs and academic campuses to becoming one of the most important drivers for the global economy, the market for compute has evolved just as quickly and is in desperate need of a globally accepted pricing mechanism and risk management tool,” stated Trabue Bland, ICE senior vice president, Futures Exchanges. “Ornn’s index, which is bringing greater transparency into the volatile cost of GPUs, is a natural fit for futures markets, and we’re excited to offer a new tool for price discovery and risk management.”
On July 1, ICE joined with compute exchange NATIVX in plans for futures based on NATIVX’s COIL Index, which tracks the price of tokenized, energy-normalized compute and connectivity. “The new contracts will offer price discovery for customers globally through a hedgeable index that will benefit from trading alongside ICE’s natural gas and power futures contracts,” Bland said.
ICE, which owns the New York Stock Exchange, started out as an electricity exchange, founder, Chair and CEO Jeffrey Sprecher recalled on Bloomberg News. “So it’s in our DNA.” Compute today “is electricity, and possibly chips, on the input. And on the output” are tokens, which are “basically the way you pay for compute.
“Somewhere in that amalgamation I suspect is going to be trading and risk transfer, whether it’s buying the inputs to the power plant and data center, or whether it’s buying the output, how the models work.”
How Goldman Sachs Research breaks down $1.019 trillion in estimated AI investments this year (total global AI-exposed hyperscaler capex).
The Commodity Futures Trading Commission has empaneled several advisory committees of industry representatives – such as Agricultural, Energy and Environmental, and Market Risk. The newest, the Innovation Advisory Committee (IAC), had the distinction of being hosted by President Trump at the White House on August 19, the day before the IAC’s inaugural meeting.
Much of the IAC membership – it includes CEOs of exchanges (ICE, Nasdaq, CME, Cboe), crypto platforms (Coinbase, Gemini, Kraken), and market infrastructures both traditional (DTCC, Options Clearing Corp.) and digital-asset (BitGo, Chainlink) – attended both events. The gatherings underscored administration support of the crypto industry and of the delayed market-structure legislation known as the CLARITY Act, as well as CFTC Chairman Michael Selig’s designation of innovation as a strategic priority.
Selig, appointed by Trump and sworn in last December, devoted a portion of his opening remarks to a “roadmap for the crypto capital of the world.” He vowed that “if CLARITY [scheduled for a September 15 Senate vote] continues to stall because of Democratic obstruction, the CFTC will utilize its existing authorities to begin establishing a regime for crypto asset markets.”
CFTC Chief Michael S. Selig
The IAC’s August 20 meeting was divided into three sessions – “crypto’s regulatory evolution,” artificial intelligence, and prediction markets (Polymarket CEO Shayne Coplan and Kalshi co-founder Luana Lopes Lara were in attendance). Compute futures found its way into the conversation.
“If the lessons of history are any guide,” said Selig, “well-functioning spot, forward, and derivatives markets emerge in conjunction with demand for scarce and economically significant commodities. Compute is proving to be no different.”
Peter C. Earle of the American Institute for Economic Research explored the potential in a July paper, “The Financialization of Compute: Forward Markets, Futures, and the Economics of Computational Capacity.”
The CFTC opened a public comment period on compute markets, which, Selig said, could transform “a costly and unpredictable input into a commodity with reliable price discovery and effective hedging, which will strengthen America’s capacity to lead the AI revolution.
“Just as American markets helped establish the gold standard for trading the commodities that powered the industrial economy,” Selig continued, “we will do the same for the commodity that will power the intelligence economy.”
The longtime trading and technology entrepreneur Don Wilson, founder and CEO of Chicago-based DRW, chimed in: “We are in a race for AI supremacy with China. Risk management instruments in the compute space are essential to reducing the cost of capital, which in turn will give the U.S. a competitive advantage.”
DRW founder Don Wilson
DRW is an investor in Silicon Data, which, according to Wilson, “has successfully created indices which are robust, well-constructed and not readily susceptible to manipulation. They leverage data from numerous neoclouds as well as from its sister company Compute Exchange” – of which Wilson is a co-founder along with tech investor Suna Said and Carmen Li, who is CEO of both Silicon Data and Compute Exchange. “They do not rely on a small number of hyperscalers’ pricing.”
Wilson advocated a fast track for compute: “It is a mistake to delay the launch of these futures by subjecting them to a lengthy comment period.”
IAC member Tushar Jain, co-founder and managing partner, Multicoin Capital, observed, “We are seeing a lot of new entrepreneurial ideas on how to structure compute derivatives. They don’t look a lot like traditional commodities.” He suggested that if granted “an innovation exemption,” the ideas could be “tested out on a smaller scale before committing the capital to scale it up and do[ing] so in a fully compliant manner.”
In US Regulatory Intelligence, Ty Greenberg, chief revenue officer and a co-founder of legal and compliance infrastructure provider Caffrey.AI, contrasted compute with futures markets in which pricing is disciplined by “the possibility of physical delivery”:
“Compute is harder. There is no single, readily observable ‘price of compute,’ and compute itself is heterogeneous across chips, regions, providers, and commitment terms. There is also no deep, transparent secondary market of the kind that exists for many commodities that are the subject of futures trading.”
What’s more, “the compute market is dominated by a relatively small number of large providers. Unlike in retail prediction markets, there is no readily observable event that establishes winners and losers. And there are no mandatory reporting requirements for physical compute transactions, so there is no public dataset from which to establish an index.”
Indicating Chinese intentions, the Shanghai municipal government included compute, electricity, and liquefied natural gas futures in a statement of its aspirations to be a “global wealth management hub,” the South China Morning Post reported.
Others are entering the fray:
-- After acquiring IMX Health, a Designated Contract Market under CFTC oversight, and having partnered with Ornn, Architect Financial Technologies said it would launch American Innovation Exchange. It described “The AI Exchange” as “the first CFTC-regulated exchange for futures and options trading on compute costs tied to multiple GPU vendors and models,” and “the definitive U.S. market for data centers, neoclouds, hyperscalers, chip manufacturers, AI model companies, and lenders, matching the rapid pace of AI industry development.”
-- BGC Group in June launched BGC Compute Infrastructure Markets (BGC CIM), a division within Energy, Commodities and Shipping “designed to bring institutional-grade market structure to one of the fastest emerging areas of the digital economy – enabling transparent price discovery, real-time risk management and more efficient execution for market participants.” Zach Espinosa, division co-head with Marc Kuber, said, “The secondary market for compute is quickly becoming one of the most consequential commodity markets in the world economy. BGC CIM has been created to meet our clients’ needs to manage their infrastructure price risk via trusted access to liquidity, market intelligence and seamless execution.”
-- Kalshi, the prediction market company that calls itself “the next-generation financial exchange,” in July introduced forward price curves for Nvidia B200, H200 and A100 chips. Although the curves are not tradable assets, “they are useful as a reference price for structuring swaps and over-the-counter compute deals,” Kalshi explained. Yet “parties looking to lock in prices or offload risk related to compute can also do so on Kalshi, via trading on the underlying markets at Kalshi’s exchange” or via its block trade functionality.
“Like every commodity before it, [compute] needs a real derivatives market,” said Kalshi CEO Tarek Mansour. “Demand for AI is only going to increase. Kalshi intends to be the exchange where all future buyers and sellers manage their risk.”
Silicon Data Founder and CEO Carmen Li
Silicon Data unveiled a GPU forward curve in April, closing a “term structure pricing” gap, with CEO Carmen Li stating, “For the first time, a CTO and a CFO can look at the same chart to decide whether to build, buy, or hedge their compute needs. What’s been missing is a way to understand not just what compute costs today, but what the market expects it to cost in the future. The forward curve begins to make that visible.”
Last December, along with an update to its flagship A100 and H100 Rental Indices, Silicon Data launched “the world’s first B200 On-Demand Rental Index, providing unprecedented transparency into the next generation of AI compute economics . . . With more than 500 enterprise, financial, and semiconductor clients globally, Silicon Data’s indices have become the trusted benchmark for measuring GPU pricing, useful life, and infrastructure cost dynamics. These updates mark one of the most substantial expansions of Silicon Data’s coverage since its inception.”
Venture capital has helped fuel the compute trading and data initiatives.
When Compute Exchange debuted early last year, it had backing from Don Wilson’s DRW and from Suna Said; the latter co-founder brought additional “street cred” from her founding of the Woodside AI venture studio.
This July, Compute Exchange rolled out a secondary marketplace to serve the growing demand for older Nvidia GPUs.
"Not every workload requires the newest generation of GPUs," said CEO Li. "For many organizations, used and refurbished hardware offers the fastest and most economical path to expanding AI infrastructure. We see the secondary GPU market as a natural step toward making AI infrastructure more efficient, transparent, and accessible.”
On August 11, nearly a year and a half since Silicon Data’s $4.7 million seed funding and less than two months ahead of the planned CME/NYMEX contracts utilizing Silicon Data benchmarks, the company announced a $30.5 million initial close of its Series A round. It was led by the Valor Atreides AI Fund. CME Ventures, DRW, F-Prime, Samsung Next, VanEck, Further, Jump, Tectonic, and Wintermute also invested. Breed, Hack, Blank VC, Sancus Ventures and SoGal Ventures participated.
“Silicon Data spent two years building the pricing and performance history a futures market needed before one existed,” said the press release. “This round funds the next phase: making sure every part of the compute economy – buyers, sellers, exchanges, and the institutions now underwriting it – builds on the same reference points.”
Ornn’s $5.7 million seed round, in October 2025, was led by Crucible Ventures and Vine Ventures. Said Crucible Capital founder Meltem Demirors, “The AI data center boom is the largest infrastructure build in human history, with $4 trillion booked to be spent by 2030. This immense capital deployment requires financial engineering to price, manage, and transfer risk between market participants. As early investors who funded and shaped market structure for digital assets, we are excited to back Ornn as they establish a new market structure and financial contracts for the AI economy – starting with tradable OTC and listed markets on compute indices.”
A $33 million funding for Ornn this June, also said to be seed stage, was led by Andreessen Horowitz, with investors including Galaxy Ventures, Nordstar and SV Angel.