Argonne LLM-Inference-Bench: SambaNova SN40L x8, SambaFlow, Mistral-7B BF16, 128in/128out, batch 16
6967.6 tok/s aggregate across 16 batched requests on 8x SN40L
Fields
id
benchmarks/argonne-sn40l-sambaflow-mistral-7b-8x-b16-128io-tpsMetric
tps_aggregate
Value
6,967.6
Unit
tok/s aggregate across 16 batched requests on 8x SN40L
Model
mistralai/Mistral-7B-v0.1
Methodology
THIRD-PARTY MEASURED (Argonne LLM-Inference-Bench, arXiv:2411.00136, All_results.csv). FP16/BF16 weights, 128 in / 128 out. Throughput = batch*(in+out)/end-to-end latency = AGGREGATE over the batch, NOT per-request decode tok/s. Batch 16 (sambaflow row max). Latency 0.588s.
Measured by
third_party
Reproducible
no