반도체·하드웨어 · 리포트
MLCommons released MLPerf Training v6.1 results, with new benchmarks focused on recommender systems and post-training for agentic systems, and record participation that opens new head-to-head comparisons.
The new recommender and agentic post-training tests give buyers a fresh, standardized basis for comparing accelerator and system performance on workloads that drive datacenter demand.
업계 전문지Slicast · 2026년 11월 16일 17:40 UTC · 글로벌 · 출처: HPCwire
중요도 55