General Compute

General Compute is an AI inference cloud company, positioning itself as the neocloud for alternative chips. The company productionizes purpose-built AI silicon from vendors such as SambaNova, Cerebras, Positron, and d-Matrix by purchasing the racks, securing data center space, and operating the hardware for customers. Its architecture splits prefill on GPUs and decode on specialized accelerators, generating tokens 5-7x faster than GPU-based competitors, serving frontier labs, fast-growing AI application companies, and asset-light clouds.