sglang/ GB200 perf

dsr1-fp4-1k1k-mid-curve

concurrency 512 · 17 days ago

cron
passed
re-run
Commit

Fix Req array token-id concatenation (#26182)

mmangkad·17 days ago
PR #26182 · Fix Req array token-id concatenation
Run
GitHub Actions27081373586-1
Slurm job5126
GPUs48·prefill 16 / decode 32
ISL / OSL1024 / 1024

Metrics

28 captured

best_of

1.00

burstiness

1.00

completed

5,120

duration

154

max_concurrency

512

mean_e2el_ms

14,821ms

mean_itl_ms

735ms

mean_tpot_ms

14.99ms

mean_ttft_ms

1,005ms

median_e2el_ms

14,760ms

median_itl_ms

719ms

median_tpot_ms

15.13ms

median_ttft_ms

736ms

num_prompts

5,120

output_throughput

30,669tok/s

p99_e2el_ms

18,782ms

p99_itl_ms

1,204ms

p99_tpot_ms

15.74ms

p99_ttft_ms

3,762ms

peak_output_tokens_per_s

38,271s

request_throughput

33.25tok/s

std_e2el_ms

1,275ms

std_itl_ms

178ms

std_tpot_ms

0.76ms

std_ttft_ms

920ms

total_input_tokens

4,717,859

total_output_tokens

4,722,209

total_token_throughput

61,309tok/s