sglang/ GB200 perf

dsr1-fp4-1k1k-mid-curve

concurrency 2,048 · 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

20,480

duration

215

max_concurrency

2,048

mean_e2el_ms

20,718ms

mean_itl_ms

1,008ms

mean_tpot_ms

20.58ms

mean_ttft_ms

1,754ms

median_e2el_ms

20,294ms

median_itl_ms

1,041ms

median_tpot_ms

21.01ms

median_ttft_ms

821ms

num_prompts

20,480

output_throughput

87,998tok/s

p99_e2el_ms

32,629ms

p99_itl_ms

1,676ms

p99_tpot_ms

21.87ms

p99_ttft_ms

13,329ms

peak_output_tokens_per_s

108,279s

request_throughput

95.41tok/s

std_e2el_ms

3,005ms

std_itl_ms

210ms

std_tpot_ms

1.37ms

std_ttft_ms

2,758ms

total_input_tokens

18,880,692

total_output_tokens

18,888,974

total_token_throughput

175,957tok/s