Intercontinental Exchange (ICE) and CME Group are developing compute futures contracts tied to the cost of GPU computing power [1, 2].
These financial instruments aim to transform AI compute into a tradable commodity similar to oil. This shift provides investors and companies a mechanism to hedge against the rapidly rising costs of the processing power required to run artificial intelligence [1, 4].
Carmen Li, CEO of Compute Exchange and Silicon Data, is involved in the development of these markets [1]. Li said the ambition behind these contracts is to create a market that could eventually exceed the size of the global oil market [1, 2].
According to reports, these compute futures contracts are expected to launch later in 2024 [1, 2]. The move reflects a growing need for price stability in the AI sector, where the demand for high-end GPUs often outstrips supply.
The push for compute derivatives is not limited to the U.S. China is also designing its own AI-token futures market [3, 5]. Reports said that China was working on this initiative as of May 28, 2024 [5], suggesting a global race to establish the primary financial infrastructure for AI resources.
By treating compute as a commodity, exchanges allow buyers to lock in prices for future capacity. This protects them from volatility in the hardware market, and the pricing whims of cloud providers [1, 4].
“The ambition behind these contracts is to create a market that could eventually exceed the size of the global oil market.”
The financialization of compute power signals a transition of AI from a purely technological race to a macroeconomic one. By creating a liquid market for GPU cycles, the industry is acknowledging that compute is now a fundamental raw material of the modern economy. The parallel efforts in the U.S. and China suggest that control over the pricing and distribution of AI compute will be a key pillar of geopolitical and economic influence.
