Bernstein: Crypto World Surpassed Traditional Exchanges in AI by Years

Derivative instruments familiar from crypto markets have stepped in much earlier than traditional exchanges to manage AI computing power costs, initiating a new era in GPU pricing.

As computing power (compute) becomes one of the most valuable commodities in the modern economy within the artificial intelligence (AI) world, crypto-native financial instruments are coming to the fore to manage price risks in this field. According to a recent report published by Bernstein, perpetual futures contracts and prediction markets have begun to be actively used in the GPU market years ahead of the official plans of major exchanges.

Bermuda-based Architect’s AX exchange has already started listing perpetual futures for GPU processing power. On the other hand, the US-regulated prediction market Kalshi is creating forward curves by offering event contracts based on rental prices for powerful NVIDIA chips such as the B200, H200, and A100. This situation demonstrates technological movement at a speed far ahead of the cash-settled futures targeted by traditional giants like CME and ICE for late 2026.

Is GPU Computing Power the New Generation of Electricity?

The perpetual contract structure, which evolved in crypto markets and has no expiration date, is being integrated into the AI world by tracking the spot index through the funding rate mechanism. Analysts compare AI computing power to electricity rather than oil; because an unused GPU hour cannot be stored and vanishes instantly. This digital brainpower capacity causes prices to rise rapidly during periods of limited supply, while derivative instruments provide protection against this volatility.

Challenges in creating liquidity and a reliable index are seen as the biggest obstacles in this market, which is still in its early stages. While startups like Silicon Data and Ornn are attempting to create tradable assets by normalizing thousands of data points, it is noted that market volume is currently influenced by speculative flows. Bernstein analysts, while closely monitoring this financial productization process at the intersection of AI and crypto, note that they also hold long positions in various crypto assets.