Intel Arc Pro B60 24GB x4, MLPerf v5.1 Server: Llama-2-70B aggregate tok/s
2110.05 output tok/s AGGREGATE across 4x Arc Pro B60 (Server; ~528 tok/s per card)
Fields
id
benchmarks/prob60dual-4x-mlperf-v5-1-llama2-70b-server-tokensMetric
tps_aggregate
Value
2,110.05
Unit
output tok/s AGGREGATE across 4x Arc Pro B60 (Server; ~528 tok/s per card)
Model
Llama-2-70B (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-2-70B, Server. 'Tokens per second' = 2110.05 verbatim from mlperf_log_summary.txt. AGGREGATE across 4 cards — ~528/card.
Measured by
vendor
Reproducible
yes
notes
Server scenario, TTFT ~161 ms min. 8 B60 dies, ~264 tok/s per die.
Notes
Server scenario, TTFT ~161 ms min. 8 B60 dies, ~264 tok/s per die.