The buildout of AI infrastructure shows no signs of slowing: companies are spending hundreds of billions of dollars a year on data centers and GPUs. As a result, compute has become the single largest cost for organizations developing AI products. Despite that scale of spending, there is not yet a straightforward, widely accepted way to price compute or to hedge exposure when compute prices move.
The company and its funding
Silicon Data has closed a $30 million Series A round to address that gap. The startup plans to establish a reference price for GPU rentals and to build an index that could serve as the settlement benchmark for Wall Street futures contracts.
The company intends to launch compute futures trading on the CME (Chicago Mercantile Exchange) on October 5, pending regulatory approval.
Why it matters
A transparent reference price and a liquid futures market for compute could allow firms to manage cost volatility and plan investments more reliably. Standardized pricing and the ability to hedge would reduce some of the financial uncertainty facing companies that rely heavily on GPUs and large-scale data-center capacity.
Discussion on TechCrunch’s Equity podcast
On an episode of the TechCrunch Equity podcast, Rebecca Bellan interviewed Steve Hou, head of research at Silicon Data. They discussed the health of the AI buildout and examined why the underlying data sometimes contradicts more pessimistic headlines about depreciating chips and stalled data center projects.
Next steps and availability
Regulatory approval is required before the CME futures can begin trading; if approved, the planned start date is October 5. For listeners who want more context, the Equity podcast is available on YouTube, Apple Podcasts, Overcast, and Spotify, and the show can be followed on X and Threads at @EquityPod.
