As the global artificial intelligence (AI) boom drives graphic processing unit (GPU) prices to unprecedented levels, Goldman Sachs and JP Morgan are reportedly exploring the launch of GPU-linked futures products.
On June 9, U.S. fintech publication PYMNTS cited The Information, an American IT media outlet, stating that the two banks are considering futures contracts tied to GPU rental prices.
According to the report, the banks are examining various ways to trade not only GPU rental costs but also computing expenses. However, discussions are still in the early stages, and there is a possibility that they may not lead to actual product launches or market openings.
The competition to develop AI models has intensified recently, making the computing costs associated with high-performance GPUs, such as Nvidia's Blackwell, a significant expense for major tech and AI companies. As a result, interest in financial products that can help manage the volatility of GPU purchase and rental prices is growing.
While Nvidia's GPU prices are not officially disclosed, the H100 is estimated to range from $25,000 to $40,000, and the B200 is reported to be between $30,000 and $50,000. Additionally, GPU rental brokerage site VAST.AI indicates that the average rental price per hour for the H100 SXM is $2.33, while the B200 averages $4.22.
PYMNTS noted that if an official market based on GPU rental prices is established, companies could manage computing cost volatility more systematically. However, challenges remain regarding how to establish reliable price benchmarks and whether related financial products can pass regulatory scrutiny.
This concept is already showing early trading examples. PYMNTS reported that inter-institutional trades have occurred based on the Ornn Compute Price Index, which tracks Nvidia's H100 GPU rental rates. This is seen as an initial attempt to index AI computing costs for actual trading purposes.
However, experts point out that for computing power to be traded as a financial product like oil or electricity, a reliable price index and sufficient trading volume must be established. PYMNTS believes that the emergence of additional price indices, participation from exchanges and clearinghouses, and regulatory decisions will be crucial in the future.
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.