sglang/ GB200 perf

dsr1-fp4-1k1k-mid-curve

concurrency 512 · 5 days ago

cron
passed
re-run
Commit

[AMD] Document Mori XGMI for Single-Node PD Disaggregation (#25094)

clintg6·5 days ago
PR #25094 · [AMD] Document Mori XGMI for Single-Node PD Disaggregation
Run
GitHub Actions27803336065-1
Slurm job5322
GPUs48·prefill 16 / decode 32
ISL / OSL1024 / 1024

Metrics

28 captured

best_of

1.00

burstiness

1.00

completed

5,120

duration

165

max_concurrency

512

mean_e2el_ms

15,834ms

mean_itl_ms

789ms

mean_tpot_ms

16.09ms

mean_ttft_ms

1,007ms

median_e2el_ms

15,770ms

median_itl_ms

782ms

median_tpot_ms

16.26ms

median_ttft_ms

741ms

num_prompts

5,120

output_throughput

28,662tok/s

p99_e2el_ms

19,462ms

p99_itl_ms

1,135ms

p99_tpot_ms

16.92ms

p99_ttft_ms

3,988ms

peak_output_tokens_per_s

35,536s

request_throughput

31.08tok/s

std_e2el_ms

1,286ms

std_itl_ms

171ms

std_tpot_ms

0.80ms

std_ttft_ms

950ms

total_input_tokens

4,717,859

total_output_tokens

4,722,209

total_token_throughput

57,297tok/s