Jump Trading and DRW are stepping up efforts to build out a financial derivatives market tied to AI compute, according to Bloomberg. The firms plan to trade compute futures and other contracts aimed at helping companies hedge volatility in GPU rental prices.
Both companies entered crypto relatively early. They are now trying to apply similar market-making experience to AI compute.
How the two firms are positioning
Jump has assembled a dedicated trading team and will use its own balance sheet to take trading risk. It also plans to make markets in compute futures being prepared by CME and Silicon Data, while monitoring related products from Intercontinental Exchange, or ICE, and Ornn.
DRW plans to take part in compute trading through three teams and has already invested in Silicon Data and Ornn. Bloomberg reported that CME’s related futures product is still awaiting regulatory approval.
The timing comes as AI infrastructure spending faces scrutiny
The expansion comes at a moment when investment in AI infrastructure is being questioned. Stocks and bonds of compute providers such as CoreWeave have come under pressure, and the market is reassessing whether AI revenue can support the cost of building data centers.
Boston Consulting Group estimates that the global AI compute market could reach $2.3 trillion by 2030.
Why the market is still early
Even so, the market for compute derivatives remains at an early stage, with limited price transparency. GPU rental prices vary widely across regions, hardware configurations and providers, which could leave futures indexes out of sync with what companies actually pay.
There is also a structural constraint: compute cannot be stored the way oil or copper can, and unused GPU time cannot be saved for future use. Whether Jump and DRW can help turn compute into a standardized asset that can be traded and hedged is still an open question.