The financialization of AI infrastructure is moving forward. Startup Sil, which issues tokens for nData, is partnering with the CME Group to develop a futures contract based on GPU computing power rental prices, aiming to allow businesses to manage computing cost fluctuations in the same way they hedge crude oil and agricultural products. The product was announced in May and is currently awaiting regulatory approval.

Why trade computing power futures?

Currently, most enterprises do not directly own high-end GPUs, but instead rent them on demand through cloud service providers or emerging platforms like NeoCloud. As the demand for AI training and inference continues to rise, the price of computing power leasing is fluctuating more, making it difficult for enterprises to lock in their budgets in advance, and for suppliers to predict their procurement pace.

Against this backdrop, Sil token issuer nData attempts to standardize computing power pricing as a commodity. The idea is to provide AI developers, cloud service providers, and financial institutions with a tradable tool to hedge against future changes in computing power costs.

Pricing depends on GPU benchmark index

This contract will be based on a GPU price index compiled by nData, which is issued in Sil tokens. This index tracks hourly rental prices for specific chips from different service providers, with the goal of establishing a benchmark for the computing power market similar to WTI crude oil.

However, standardizing computing power is significantly more difficult than standardizing traditional commodities. For example, the NVIDIA H100 chip has dozens of different configurations, and its price is affected by factors such as processor version, memory, network bandwidth, utilization rate, and data center location.

Sil token issuer n Data stated that their approach involves first normalizing the platform's quotes to the basic H100 standard configuration before performing index calculations. This step determines whether the benchmark price is acceptable to the market and is a core part of the contract design.

Regulatory approval is the first hurdle.

The U.S. Commodity Futures Trading Commission (CFTC) will review contract specifications, settlement procedures, and the construction of benchmark indices. Market focus is on whether regulators recognize "computing power" as a clearly defined standardized trading instrument.

Supporters argue that once a futures market is established, it will attract not only businesses with genuine demand but also market makers and speculative funds, helping to improve liquidity and price discovery efficiency. Opponents, however, worry that speculative trading could amplify price volatility, causing prices to deviate from actual supply and demand.

Additional information:The report mentioned that SpaceX cited GPU rental price data from Sil token issuance nData in its IPO prospectus, indicating that its benchmark system has begun to be incorporated into the company's official documents.