Intel Arc Pro B60 24GB x4, MLPerf v5.1 Offline: Llama-3.1-8B aggregate tok/s
6472.37 output tok/s AGGREGATE across 4x Arc Pro B60 (Offline; ~1618 tok/s per card)
Fields
id
benchmarks/prob60dual-4x-mlperf-v5-1-llama31-8b-offline-tokensMetric
tps_aggregate
Value
6,472.37
Unit
output tok/s AGGREGATE across 4x Arc Pro B60 (Offline; ~1618 tok/s per card)
Model
Llama-3.1-8B (FP8)
Methodology
VENDOR-MEASURED (MLPerf v5.1 closed submission, Intel). 4x Arc Pro B60 24GB GDDR6 (PCIe Gen5 x8) on Xeon 698X host, PyTorch vLLM, Llama-3.1-8B, Offline. 'Tokens per second' = 6472.37 verbatim from mlperf_log_summary.txt. AGGREGATE across 4 cards — ~1618/card.
Measured by
vendor
Reproducible
yes
notes
4x dual-GPU card = 8 B60 dies, ~809 tok/s per die.
Notes
4x dual-GPU card = 8 B60 dies, ~809 tok/s per die.