Andromeda Cluster

Andromeda Cluster operates a global GPU compute marketplace that connects AI teams and researchers with high-performance GPU infrastructure across dozens of providers. The platform handles the full transaction: sourcing capacity, benchmarking hardware, pricing, certifying it and deploying clusters. Over a billion GPU-hours have flowed through the network, and the company began with a single multi-thousand-GPU cluster.

Founded in 2022 by Nat Friedman and Daniel Gross, the company was initially built to serve the compute needs of their venture portfolio. Its stated aim is to become the liquidity layer for global AI compute - analogous, in its founders' framing, to the electrical grid for manufacturing - so that researchers and engineers can build rather than negotiate hardware contracts.

The technical work spans GPU compute infrastructure, AI and ML training infrastructure, cloud marketplace mechanics, hardware benchmarking, compute pricing and certification, and distributed cluster deployment. A core part of the mandate is building the benchmarks, standards and trust mechanisms on which efficient compute transactions depend. The company serves AI startups, AI researchers and teams training ML models, and operates globally.