Silicon Data and CME Prepare Futures on AI Computing Power Rentals
Silicon Data and CME plan futures tied to Nvidia GPU rental prices, but regulator CFTC extended its review and delayed the launch.
- published on Newsmetry
- original source published
the public page opens no earlier than 15 minutes after we detect the story. approved stories may reach the api sooner. connect the api →
what happened
According to WSJ, Silicon Data, founded by Carmen Li, is working with CME Group to prepare futures on rentals of computing power for AI. The contracts will be based on indexes that track what Nvidia chips cost to rent. The launch, originally set for October 5, was postponed after the CFTC extended its review by another 45 days.
CME intends to list two contracts, each covering one month of rent: one for the H100 and one for the B200. Settlement will be in cash, so no compute capacity has to be delivered. Ornn, together with Intercontinental Exchange, is also preparing similar contracts.
Critics compare the idea to dot-com era bandwidth futures, which failed after a glut of fiber-optic networks. The regulator has flagged the risk of manipulation by dominant players and the lack of standardization and liquidity.
why it matters
If the contracts launch, companies could hedge one of their largest AI cost lines, and traders could take positions on the sector without buying shares. For markets, this would create a new benchmark for computing prices, though its reliability remains open to question.
market impact
AI stocks and chipmakers: a new way to bet on GPU rental prices, with no launch date yet
tags
instruments
sources
- Want to Short the AI Boom? She Wants to Give You a Way to Bet on 'Compute' Prices — WSJ, 21 October 2026 at 19:32 UTC
prepared by the newsmetry newsroom with ai assistance from the sources listed below. not investment advice.
the public page opens no earlier than 15 minutes after we detect the story. approved stories may reach the api sooner.
get stories in real time ↗