
CME Group will launch futures on Nvidia AI chip rental costs.
Contracts settle on Silicon Data benchmarks for H100 and B200.
This lets investors bet on compute prices directly.
What happened
CME Group plans to launch futures on Oct. 5 tied to the hourly rental cost of Nvidia's H100 and B200 graphics processors, pending regulatory approval. The contracts would settle against benchmarks from Silicon Data, which tracks what companies pay to rent those chips.
Why it matters
Nvidia is moving beyond simply selling chips, helping to finance customers, and participating in rental revenue as Wall Street increasingly likens Nvidia to the central bank of AI. CEO Jensen Huang put the shift more simply: "Now, compute is revenue."
What to watch
Silicon Data's H100 benchmark is currently around $2.68 per GPU-hour, while its newer B200 benchmark is about $5.66. Both have moved substantially over the past year, and not always together. A futures contract does not reserve the chips themselves; it pays out based on where the benchmark price goes.
Ask the AI about this article →
Nvidia's forecast of roughly 70% revenue growth in fiscal 2028 underscores the scale of the AI boom, but pricing that compute remains difficult. The company is not just selling chips—it is financing customers and sharing in rental revenue, which has led Wall Street to compare Nvidia to the central bank of AI. The new futures market aims to put a tradable price on compute, much like commodities.
The challenge of standardizing such contracts is not new. Past attempts, like DRAM futures, struggled with industry agreement on what the standard chip would be, while weather futures faced issues with specificity. Bandwidth, especially during the fiber boom, serves as a cautionary tale—Enron's effort to commoditize network capacity never caught on, and excess capacity later drove prices down.
For investors, these futures could offer a way to hedge or speculate on AI infrastructure costs, but they do not guarantee access to the physical chips. The success of the market will likely hinge on how well the benchmarks reflect real-world rental dynamics, which vary by provider, location, and term.
For example, today's edition would include:
AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.
Free · takes 30 seconds · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. Q&As are published on this page for other readers too.
Clearway Energy Group signed three long-term power purchase agreements (PPAs) with Alphabet's Google for nearl…

A YouGov survey of 1,250 U.S

EDM producer Max "H4RRIS" Harris and Italian turntablist-turned-producer Nihil Young are publicly calling out…

On Aug. 24, Virtuals Protocol launched a new suite of AI agent creation and ownership tools on Solana, enablin…

Meta Platforms released Muse Glimmer, a 30-billion-parameter AI model with open weights, on Aug

OpenAI is ending its contract with Cursor, effective November 12, 2026, citing Elon Musk's companies' history…
