AI Summary
5 min readThe CME Wants to Turn AI Computing Power Into a Commodity
Wall Street will soon be trading AI computing power like oil or wheat. The CME Group, one of the world's largest futures and options exchanges, announced it will launch "compute futures" — each contract representing one month's rent on an Nvidia H100 chip. The stated goal is to give data center providers and AI companies price transparency and the ability to hedge against swings in computing costs. But the move raises two uncomfortable questions: Is compute actually a commodity? And should we be trading it at all?
Why Compute Isn't Like Oil
Rohan Goswami, business reporter at Semaphore, explains that the idea has been in the works for months. Larry Fink made an oblique reference to it at the Milken conference earlier this year, and data providers have been building the infrastructure for both the CME and ICE to offer these products. The exchanges like to compare compute to oil or electricity — scarce, hard to transport, diffuse. But Goswami pushes back: "Oil is actually finite." Compute, by contrast, depends entirely on what one player decides to do: Nvidia.
Continue reading the full summary in the app — free to try.
Read Full Summary →Free • No credit card required
Never miss an episode of Prof G Markets
Get every new episode summarized in your inbox — free, ~5 minutes to read.
No spam. Unsubscribe anytime.
What you'll learn
- 1 (02:03) **Episode Introduction & Market Vitals** - Ed Elson recaps the previous day's market data: S&P 500 rising, odds of a rate hike falling, and a 19% rally in CoreWeave shares.
- 2 (02:50) **The Announcement: Compute Futures as a Tradable Commodity** - The CME announced it will launch futures contracts representing one month's rent on an Nvidia H100 chip, aiming to turn computing power into a standardized commodity.
- 3 (03:32) **Interview Begins: Rohan Goswami on the Origins of Compute Futures** - Business reporter Rohan Goswami explains that this idea has been in the works for months, with Larry Fink referencing it earlier in the year.
- 4 (04:33) **The Nvidia Monopoly Problem** - The core issue is that Nvidia, as the sole dominant chip maker, controls the entire forward price curve of compute.
- 5 (05:58) **The Fungibility Problem: Is Compute a Real Commodity?** - The debate over whether AI compute can be treated like oil, gold, or wheat hinges on the fact that chips are not fungible—they depreciate and vary by generation.
- 6 (08:21) **The $500 Billion Financing Package: Spreading the Risk** - The CME's move is linked to Nvidia's massive financing MOU with Wall Street giants (KKR, Blackstone, Goldman Sachs), which aims to legitimize Nvidia compute as collateral for debt.
- 7 (10:30) **Circular Financing Concerns: Does This Solve Them?** - Jensen Huang claims this MOU should assuage fears about circular financing, where AI companies invest in each other and book recycled money as revenue.
+ Full timestamped outline available in the app
Guests on this episode
Show Notes
Ed Elson is joined by Rohan Goswami to break down the CME’s move to offer AI compute futures such that computing power can be traded like a commodity. Then, Mark Zandi returns to unpack the latest inflation data from the consumer price index. Finally, Ed shares his thoughts on what the recent high profile departures from OpenAI mean for the markets.
Rohan Goswami is a business reporter at Semafor and the co-host of Compound Interest. Mark Zandi is the chief economist at Moody’s Analytics.
Subscribe to the Prof G Markets Youtube Channel
Check out our latest Prof G Markets newsletter
Follow Prof G Markets on Instagram
Follow Ed on Instagram, X and Substack
Follow Scott on Instagram
Send us your questions or comments by emailing [email protected]
Learn more about your ad choices. Visit podcastchoices.com/adchoices
More from this podcast
Prof G Markets →