Cerebras

A hardware company building wafer scale processors and a hosted inference service for large models. Readers can learn about the wafer scale engine design, inference throughput claims and benchmarks, how the chips compare with GPU clusters, funding and customers, and developer access to the API.