This Equity podcast episode introduces Silicon Data, a startup that just closed a $30 million Series A with a specific ambition: to become the reference price for GPU rental and the settlement index for a Wall Street futures contract. The premise is that, despite hundreds of billions of dollars flowing annually into data centers and GPUs, compute has become the single largest cost for AI product builders, yet there is still no straightforward way to price it or to hedge exposure when prices move. Silicon Data is trying to fill that gap by establishing a standard market price, and the company plans to launch compute futures trading on the CME on October 5th, pending regulatory approval.

The episode features host Rebecca Bellan speaking with Steve Hou, head of research at Silicon Data, about the broader state of the AI buildout. According to the episode description, the discussion centers on why the underlying data tells a different story than headline narratives about depreciating chips and stalled data centers. The framing suggests that Silicon Data‘s research challenges the more pessimistic read of the infrastructure cycle, though specifics of that data are not detailed in the description.

This is a brief episode write-up rather than a full transcript, so the substantive content is limited to the startup’s business model, its funding news, and the planned CME launch date. The core value proposition is that compute is the biggest cost in AI development but lacks a transparent market mechanism, and Silicon Data aims to create that mechanism by indexing GPU rental prices and enabling futures-based hedging. The regulatory approval caveat and the specific October 5th launch target are the key concrete details offered.