sglang/ GB200 perf

dsr1-fp8-1k1k-max-tpt

concurrency 4,096 · 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

40,960

duration

351

max_concurrency

4,096

mean_e2el_ms

33,681ms

mean_itl_ms

1,631ms

mean_tpot_ms

33.41ms

mean_ttft_ms

2,927ms

median_e2el_ms

32,731ms

median_itl_ms

1,655ms

median_tpot_ms

33.78ms

median_ttft_ms

1,303ms

num_prompts

40,960

output_throughput

107,646tok/s

p99_e2el_ms

54,346ms

p99_itl_ms

2,715ms

p99_tpot_ms

35.56ms

p99_ttft_ms

23,174ms

peak_output_tokens_per_s

131,051s

request_throughput

117tok/s

std_e2el_ms

5,213ms

std_itl_ms

376ms

std_tpot_ms

1.72ms

std_ttft_ms

4,843ms

total_input_tokens

37,769,666

total_output_tokens

37,742,239

total_token_throughput

215,371tok/s