MLPerf v6.1 Dell R770 2x Xeon 6787P: llama3.1-8b INT8 Server aggregate tok/s
516.012 tok/s AGGREGATE over 2x Xeon 6787P (Server scenario)
Fields
id
benchmarks/mlperf-v6-1-dell-2x6787p-llama31-8b-int8-serverMetric
tps_aggregate
Value
516.012
Unit
tok/s AGGREGATE over 2x Xeon 6787P (Server scenario)
Model
llama3.1-8b
Methodology
MLPerf Inference v6.1 closed division, CPU-only (summary.csv accelerator_model_name is N/A). Values copied from summary.csv, not from a per-system mlperf_log_summary.txt. AGGREGATE throughput for the whole node, NOT per-request decode and NOT per-socket. Prompt length, output length, and batch size are not in the summary row. weight_data_types INT8. Framework column is PyTorch; the Dell notes field additionally names vLLM with an FP8 KV cache. Scenario: Server. System: Dell PowerEdge R770, 2x Xeon 6787P, 16x64GB DDR5-6400. Server scenario companion. vLLM, FP8 KV cache, per the summary.csv notes. 516.012 tok/s aggregate across both sockets. measured_by vendor: these are submitter results inside the MLCommons closed division (Intel, Dell, Supermicro, Quanta, Red Hat, Cisco), not an independent re-run.
Measured by
vendor
Reproducible
yes
notes
Server scenario companion. vLLM, FP8 KV cache, per the summary.csv notes. 516.012 tok/s aggregate across both sockets.
Notes
Server scenario companion. vLLM, FP8 KV cache, per the summary.csv notes. 516.012 tok/s aggregate across both sockets.