MLPerf v6.1 Naeem Khoshnevis 1x H200-SXM-141GB (RDI): llama3.1-8b Offline 7796.59 tok/s
7796.59 tok/s aggregate Offline
Fields
id
benchmarks/mlperf-v6-1-1xh200-llama31-8b-offline-tokensMetric
tps_aggregate
Value
7,796.59
Unit
tok/s aggregate Offline
Model
Llama-3.1-8B FP8
Methodology
MLPerf Inference v6.1 published result (single H200-SXM-141GB, RDI availability category; software: TensorRT-LLM 1.0.0 (FP8 W8A8 trtllm-build/trtllm-serve), CUDA 12.9). System-AGGREGATE output tokens/s for the named scenario, submitter-measured under MLCommons audit. RDI (research/dev/internal) category - not an Available-systems submission. Rare single-GPU H200 row.
Measured by
vendor
Reproducible
yes
notes
ENGINE_ID NOTE: pinned to 'tensorrt-llm' as the closest existing engine record; the submission's actual stack is named in this methodology string.
Notes
ENGINE_ID NOTE: pinned to 'tensorrt-llm' as the closest existing engine record; the submission's actual stack is named in this methodology string.