MLPerf Inference v6.1 Adds First Agentic AI Benchmarks

MLCommons released MLPerf Inference v6.1 with a record 30 submitters and up to 5.7x performance gains. It introduces the first agentic inference benchmark on datacenter hardware and the first trillion-parameter model deployment.

2026-09-16 ~ 2026-09-17 · 2 related posts