sglang/ GB200 perf

dsr1-fp8-1k1k-max-tpt

concurrency 2,048 · 2 days ago

cron
passed
re-run
Commit

[AMD] Register DP attention test (#28495)

bingxche·2 days ago
PR #28495 · [AMD] Register DP attention test
Run
GitHub Actions27927465630-1
Slurm job5331
GPUs48·prefill 16 / decode 32
ISL / OSL1024 / 1024

Metrics

28 captured

best_of

1.00

burstiness

1.00

completed

20,480

duration

285

max_concurrency

2,048

mean_e2el_ms

27,302ms

mean_itl_ms

1,344ms

mean_tpot_ms

27.52ms

mean_ttft_ms

1,946ms

median_e2el_ms

26,908ms

median_itl_ms

1,376ms

median_tpot_ms

27.68ms

median_ttft_ms

1,009ms

num_prompts

20,480

output_throughput

66,328tok/s

p99_e2el_ms

37,596ms

p99_itl_ms

1,974ms

p99_tpot_ms

28.94ms

p99_ttft_ms

12,326ms

peak_output_tokens_per_s

80,793s

request_throughput

71.92tok/s

std_e2el_ms

2,886ms

std_itl_ms

246ms

std_tpot_ms

0.95ms

std_ttft_ms

2,632ms

total_input_tokens

18,880,692

total_output_tokens

18,888,974

total_token_throughput

132,628tok/s