Computing power is emerging as a tradable asset class as CME Group partners with Silicon Data to launch the first futures contracts tied to the cost of running artificial intelligence chips, according to announcements from the companies. The upcoming derivatives, slated for October 5 pending regulatory approval, will allow companies and investors to trade and hedge the price of AI computing capacity similarly to traditional commodities like oil and electricity.
CME Group and Silicon Data Target GPU Rental Pricing Transparency
For years, companies purchasing identical graphics processing unit capacity have faced wildly different pricing structures without a clear market benchmark, according to Carmen Li, CEO of Silicon Data. The new compute futures aim to solve this discrepancy by creating a public, tradable reference price for the core resource powering modern artificial intelligence systems.
The contracts will track the rental costs of Nvidia’s H100 and newer Blackwell B200 hardware. According to exchange details, pricing will rely on Silicon Data indexes that monitor hourly GPU rental rates, with each individual contract representing one month of rent for an Nvidia H100.
Financing the Massive AI Infrastructure Buildout
Wall Street continues to develop new financial instruments to gain exposure to the rapid expansion of artificial intelligence infrastructure. Alongside these derivatives, Nvidia has been collaborating with major global asset managers on initiatives aimed at channeling up to $500 billion into AI infrastructure development.
Pro Tip: Industry developers and data-center operators can utilize these upcoming compute futures to effectively hedge their operational costs or secure future revenues against market volatility, bypassing the need for direct physical investments in data centers.
These financial tools create an entirely new layer within the emerging AI financial ecosystem. Investors can gain exposure directly to the underlying computing capacity rather than purchasing physical chips, data centers, or traditional equities in the companies building them.
Frequently Asked Questions
When will the compute futures contracts launch?
The contracts are scheduled to launch on October 5, pending regulatory approval, according to CME Group and Silicon Data.
Which hardware will these futures contracts track?
The contracts track the rental costs of Nvidia H100 and newer Blackwell B200 graphics processing units, based on hourly rental price indexes provided by Silicon Data.
Who can use these compute futures?
AI developers, data-center operators, and institutional investors can use the contracts to hedge operational costs, secure revenues, or gain financial exposure to AI computing capacity.
Stay Updated on AI Infrastructure Markets
Explore our latest coverage on financial technology and market trends by subscribing to our newsletter or dropping a comment below.
Related reading