Silicon Data has raised an initial $30.5 million closing of its Series A financing to expand its independent pricing, performance and market intelligence infrastructure for the AI compute economy.
The round was led by the Valor Atreides AI Fund, with additional investment from CME Group, DRW, F-Prime, Samsung, VanEck, Further, Jump, Tectonic and Wintermute. Breed, Hack, Blank VC, Sancus Ventures and SoGal Ventures also participated.
Silicon Data plans to use the funding to expand across four primary areas: benchmark pricing, GPU performance measurement through SiliconMark, institutional market and alternative data, and risk infrastructure supporting derivatives, insurance and credit markets.
The financing comes as CME Group prepares to use Silicon Data benchmarks as the reference price for its planned cash-settled GPU futures market, subject to regulatory approval.
The planned futures product would provide market participants with a regulated instrument for managing changes in GPU rental prices using a published daily benchmark.
Silicon Data is seeking to establish standardized reference points for AI compute as spending on GPUs and related infrastructure grows. The company’s broader goal is to provide the pricing transparency and performance measurement infrastructure commonly found in mature commodity and financial markets.
Since raising a $4.7 million seed round in March 2025, Silicon Data has grown to more than 1,000 registered users spanning semiconductor manufacturers, AI model developers and financial institutions.
The company has developed nine financial-grade indices covering GPUs and large language models involved in the global AI infrastructure buildout.
Silicon Data also provides market intelligence and alternative datasets designed to help investors, infrastructure operators and financial institutions evaluate supply, demand, utilization, pricing and economics throughout the compute market.
Part of the Series A financing will support expansion of SiliconMark, the company’s system for independently measuring the real-world performance of physical GPU infrastructure.
Silicon Data noted that two computing clusters using identical GPUs can produce significantly different real-world performance based on factors including networking, topology and system configuration.
SiliconMark is designed to normalize those performance differences, which the company believes could help market participants measure and hedge compute output risk and potentially support physically delivered compute markets in the future.
Silicon Data spent approximately two years building historical pricing and performance data intended to support the development of financial markets around compute.
The latest capital will support the company’s effort to establish common reference points for compute buyers and sellers, exchanges, infrastructure providers and financial institutions underwriting AI infrastructure.
Silicon Data provides pricing indices, performance benchmarks, forward curves and institutional market data for the AI compute economy. Its products are used by compute buyers and sellers, financial institutions, exchanges and infrastructure operators.
KEY QUOTES:
“Compute is becoming one of the most important inputs in the global economy, but the infrastructure for measuring and managing it is still remarkably immature.”
“Every mature market eventually develops independent referees, benchmarks that tell participants what something is worth, whether it performs as promised, and how risk should be measured. That is the role we are building Silicon Data to play for compute.”
Carmen Li, CEO of Silicon Data
“Compute is now the seed and equipment of the AI economy, being built out at national-infrastructure scale with none of that machinery underneath it. Silicon Data is building the price layer for compute futures that will make the AI buildout easier to finance, and that’s a piece this cycle has been missing.”
Gavin Baker, Managing Partner and CIO at Atreides Management
“Great markets require trusted measurement. Silicon Data is building foundational data and market intelligence infrastructure that the compute economy needs as artificial intelligence scales.”
“Real-time GPU pricing transparency and performance benchmarking insights create durable value across cycles and uncertainty.”
Antonio Gracias, Founder, CEO and Chief Investment Officer at Valor Equity Partners
