sglang/ GB200 perf

dsr1-fp8-1k1k-max-tpt

concurrency 6,144 · 4 days ago

cron
passed
re-run
Commit

[cookbook] Laguna-M.1: add PD disaggregation section (#28737)

Jiminator·4 days ago
PR #28737 · [cookbook] Laguna-M.1: add PD disaggregation section
Run
GitHub Actions27858605416-1
Slurm job5324
GPUs48·prefill 16 / decode 32
ISL / OSL1024 / 1024

Metrics

28 captured

best_of

1.00

burstiness

1.00

completed

61,440

duration

459

max_concurrency

6,144

mean_e2el_ms

44,501ms

mean_itl_ms

1,801ms

mean_tpot_ms

36.90ms

mean_ttft_ms

10,532ms

median_e2el_ms

44,062ms

median_itl_ms

1,756ms

median_tpot_ms

37.47ms

median_ttft_ms

11,020ms

num_prompts

61,440

output_throughput

123,309tok/s

p99_e2el_ms

70,727ms

p99_itl_ms

3,409ms

p99_tpot_ms

42.43ms

p99_ttft_ms

33,090ms

peak_output_tokens_per_s

175,457s

request_throughput

134tok/s

std_e2el_ms

6,658ms

std_itl_ms

590ms

std_tpot_ms

4.76ms

std_ttft_ms

6,100ms

total_input_tokens

56,636,934

total_output_tokens

56,621,450

total_token_throughput

246,652tok/s