시작하기리더보드Decode calculator모델OrganizationsReportsMatrix runs하드웨어벤치마크마켓플레이스렌탈ProAPI 문서
언어
Actual Computer — Every computer, one endpoint

Speed test matrix

Laguna-XS-2.1-FP8 on H200 NVL 141GB

sglang · 0.5.6.post3.dev9269+g834400705 · FP86 contexts (8K–251K) × 8 concurrency levels (1–128) · prefill at 8K, 16K, 32K, 64K, 128K, 251K, 256K
by @keennaylmx dev importMeasured

Hardware

Type
Discrete GPU
GPU
H200 NVL
VRAM
141 GB
OS
linux

Engine

Engine
sglang
Version
0.5.6.post3.dev9269+g834400705
Quantization
FP8
Backend
cuda
Build
cuda-13.2
Commit
834400705f2d

Run

Served model
poolside
Revision
main
Decode test
Fixed duration
Prefill mode
Measured during decode warm-up
Time per measurement
30 s
Max output
4,096 tokens

Launch command

env SGLANG_SHARED_EXPERT_TP1=1 "$HOME/env_h200-poolside-sglang/bin/sglang" serve --model-path poolside/Laguna-XS-2.1-FP8 --served-model-name poolside --trust-remote-code --tp 1 --reasoning-parser poolside_v1 --tool-call-parser poolside_v1 --context-length 262144 --mem-fraction-static 0.850588 --enable-metrics --host 0.0.0.0 --port 8000 --api-key <redacted>

Prefill

prompt tok/s by context length
8K context41.7K197 ms TTFT
16K context41.5K391 ms TTFT
32K context37.7K855 ms TTFT
64K context30.1K2,135 ms TTFT
128K context20.9K6,131 ms TTFT
251K context13K19,283 ms TTFT
256K context12.9K19,822 ms TTFT

Sustained decode

total tok/s · rows are context length, columns are simultaneous requests
Context1248163264128
8K189.875 ms TTFT230.8102 ms TTFT605.999 ms TTFT865.2132 ms TTFT1,567.4196 ms TTFT2,409.8336 ms TTFTOver capacityOver capacity
16K186.7100 ms TTFT212.1140 ms TTFT585.2146 ms TTFT783.2186 ms TTFT1,447.2329 ms TTFT2,045.3636 ms TTFTOver capacityOver capacity
32K180.7125 ms TTFT125.1156 ms TTFT494.2213 ms TTFT666.6379 ms TTFT1,125.1680 ms TTFTOver capacityOver capacityOver capacity
64K170.6173 ms TTFT161.2290 ms TTFT450.9402 ms TTFT724.6791 ms TTFTOver capacityOver capacityOver capacityOver capacity
128K153.8292 ms TTFT222.8439 ms TTFT417.31,035 ms TTFTOver capacityOver capacityOver capacityOver capacityOver capacity
251K128.5510 ms TTFT208.8902 ms TTFTOver capacityOver capacityOver capacityOver capacityOver capacityOver capacity

Over capacity: not run because this many requests at this context would exceed the server's KV cache.

TTFT is measured after the benchmark warms the server's prompt cache, so it mostly excludes prefill.

All measurements

Latencies are medians. Total tok/s is output across all simultaneous requests.
PhaseContextRequests at onceStatusTotal tok/sPer request tok/sTTFTTime between tokensAverage active requestsCompleted requestsOutput tokensMeasured for
Prefill8K1Valid41.7K—197 ms——1——
Prefill16K1Valid41.5K—391 ms——1——
Prefill32K1Valid37.7K—855 ms——1——
Prefill64K1Valid30.1K—2,135 ms——1——
Prefill128K1Valid20.9K—6,131 ms——1——
Prefill251K1Valid13K—19,283 ms——1——
Prefill256K1Valid12.9K—19,822 ms——1——
Sustained decode8K1Valid189.8190.475 ms5.3 ms1.015,69330.0 s
Sustained decode8K2Valid230.8114.7102 ms8.7 ms2.906,92430.0 s
Sustained decode8K4Valid605.9152.399 ms6.6 ms4.0418,17630.0 s
Sustained decode8K8Valid865.2107.7132 ms9.3 ms10.8025,95230.0 s
Sustained decode8K16Valid1,567.498.2196 ms10.2 ms16.0047,02430.0 s
Sustained decode8K32Valid2,409.874.1336 ms13.5 ms33.6072,28830.0 s
Sustained decode8K64Over capacity————————
Sustained decode8K128Over capacity————————
Sustained decode16K1Valid186.7187.4100 ms5.3 ms1.015,60030.0 s
Sustained decode16K2Valid212.1100.6140 ms9.9 ms21.206,36430.0 s
Sustained decode16K4Valid585.2146.5146 ms6.8 ms4.0417,55630.0 s
Sustained decode16K8Valid783.297.8186 ms10.2 ms11.3023,49630.0 s
Sustained decode16K16Valid1,447.288.4329 ms11.3 ms16.5043,40830.0 s
Sustained decode16K32Valid2,045.363.3636 ms15.8 ms37.9061,34430.0 s
Sustained decode16K64Over capacity————————
Sustained decode16K128Over capacity————————
Sustained decode32K1Valid180.7181.6125 ms5.5 ms1.015,42130.0 s
Sustained decode32K2Valid125.161.7156 ms16.2 ms31.903,75230.0 s
Sustained decode32K4Valid494.2127.3213 ms7.9 ms4.7414,82630.0 s
Sustained decode32K8Valid666.683.6379 ms12.0 ms12.0019,99230.0 s
Sustained decode32K16Valid1,125.168.8680 ms14.5 ms20.6033,74430.0 s
Sustained decode32K32Over capacity————————
Sustained decode32K64Over capacity————————
Sustained decode32K128Over capacity————————
Sustained decode64K1Valid170.6171.8173 ms5.8 ms1.015,11730.0 s
Sustained decode64K2Valid161.278.5290 ms12.7 ms14.304,83430.0 s
Sustained decode64K4Valid450.9109.3402 ms9.1 ms4.6013,52430.0 s
Sustained decode64K8Valid724.691.3791 ms11.0 ms8.0021,73630.0 s
Sustained decode64K16Over capacity————————
Sustained decode64K32Over capacity————————
Sustained decode64K64Over capacity————————
Sustained decode64K128Over capacity————————
Sustained decode128K1Valid153.8155.4292 ms6.4 ms1.014,61430.0 s
Sustained decode128K2Valid222.8107.0439 ms9.3 ms4.406,68430.0 s
Sustained decode128K4Valid417.3102.51,035 ms9.8 ms4.0012,51630.0 s
Sustained decode128K8Over capacity————————
Sustained decode128K16Over capacity————————
Sustained decode128K32Over capacity————————
Sustained decode128K64Over capacity————————
Sustained decode128K128Over capacity————————
Sustained decode251K1Valid128.5131.0510 ms7.6 ms1.013,85430.0 s
Sustained decode251K2Valid208.8100.2902 ms10.0 ms2.206,26430.0 s
Sustained decode251K4Over capacity————————
Sustained decode251K8Over capacity————————
Sustained decode251K16Over capacity————————
Sustained decode251K32Over capacity————————
Sustained decode251K64Over capacity————————
Sustained decode251K128Over capacity————————

Over capacity: not run because this many requests at this context would exceed the server's KV cache.