CME Group CEO on Demand for AI Compute: It'll Make a Great Market Going Forward

Why It Matters

By turning AI compute into a hedgeable commodity, CME gives firms a tool to manage cost volatility, potentially accelerating AI adoption and reshaping tech‑sector financing.

Key Takeaways

• CME and Silicon Data launch futures contracts for AI compute.
• Futures aim to provide risk‑management for GPU owners and AI developers.
• Transparent pricing will reduce volatility spikes from new hardware releases.
• Contract suite can quickly add products as new GPUs hit market.
• Treating compute as a commodity mirrors traditional exchanges but is unprecedented.

Summary

CME Group chief executive Terry Duffy announced a partnership with index provider Silicon Data to create a dedicated futures market for artificial‑intelligence compute power. The new contracts will let market participants trade the cost of GPU capacity, effectively treating compute as a tradable commodity.

The move addresses a glaring gap: hyperscalers own expensive GPUs and lease them, while AI developers face unpredictable rental rates and no hedging tools. Duffy highlighted that transparent pricing and futures contracts can mitigate risk for both sides, smoothing out spikes that occur when new hardware launches.

Duffy cited the benchmark H100 rental data and emphasized Silicon Data’s ability to roll out new indices quickly—so when NVIDIA releases a next‑gen chip, a corresponding futures contract can be listed in days. He described the product suite as “nimble” and essential for a market that currently lacks formal risk‑management mechanisms.

If successful, the compute futures market could standardize pricing, attract institutional capital, and create a new asset class that mirrors traditional commodities while reflecting the unique dynamics of the AI ecosystem.

Original Description

Terry Duffy, CME Group CEO, joins 'Power Lunch' to discuss the company's new futures market for computing power.