Speed test matrix
gemma-4-31B-it-NVFP4 on RTX PRO 6000 Blackwell Server Edition 96GB
vllm · 0.30.1rc1.dev452+g73a583112 · NVFP46 contexts (8K–251K) × 8 concurrency levels (1–128) · prefill at 8K, 16K, 32K, 64K, 128K, 251K, 256K
Hardware
- Type
- Discrete GPU
- GPU
- RTX PRO 6000 Blackwell Server Edition
- VRAM
- 96 GB
- OS
- linux
Engine
- Engine
- vllm
- Version
- 0.30.1rc1.dev452+g73a583112
- Quantization
- NVFP4
- Backend
- cuda
- Repository
- github.com/vllm-project/vllm
- Build
- cuda-13.2
- Commit
73a5831127a9
Run
- Served model
- gemma
- Revision
- main
- Benchmark tool
- llm-inference-bench v0.7.6
- Decode test
- Fixed duration
- Prefill mode
- Measured during decode warm-up
- Time per measurement
- 30 s
- Max output
- 4,096 tokens
Launch command
env MAX_JOBS=2 FLASHINFER_NVCC_THREADS=1 TORCHINDUCTOR_COMPILE_THREADS=1 "$HOME/env_rtxpro6k-redhatai-vllm-73a5831127a9d2b8/bin/vllm" serve RedHatAI/gemma-4-31B-it-NVFP4 --served-model-name gemma --trust-remote-code --tensor-parallel-size 1 --reasoning-parser gemma4 --enable-auto-tool-choice --tool-call-parser gemma4 --max-model-len 262144 --gpu-memory-utilization 0.83 --host 0.0.0.0 --port 8000 --api-key <redacted> --limit-mm-per-prompt '{"image":4}' --async-schedulingPrefill
prompt tok/s by context length8K context8,1221,010 ms TTFT
16K context6,4932,504 ms TTFT
32K context4,4487,277 ms TTFT
64K context2,73823,582 ms TTFT
128K context1,57781,815 ms TTFT
251K context869.0291,002 ms TTFT
256K context853.0302,427 ms TTFT
Sustained decode
total tok/s · rows are context length, columns are simultaneous requests| Context | 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128 |
|---|---|---|---|---|---|---|---|---|
| 8K | 53.353 ms TTFT | Looped | Looped | Looped | Looped | 858.3322 ms TTFT | Over capacity | Over capacity |
| 16K | 50.176 ms TTFT | 100.2107 ms TTFT | Looped | Looped | 442.0324 ms TTFT | Over capacity | Over capacity | Over capacity |
| 32K | 46.3121 ms TTFT | Looped | Looped | Looped | Over capacity | Over capacity | Over capacity | Over capacity |
| 64K | 41.9217 ms TTFT | Looped | Looped | Over capacity | Over capacity | Over capacity | Over capacity | Over capacity |
| 128K | 35.2490 ms TTFT | 70.1642 ms TTFT | Over capacity | Over capacity | Over capacity | Over capacity | Over capacity | Over capacity |
| 251K | 26.5Requests queued1,133 ms TTFT | Over capacity | Over capacity | Over capacity | Over capacity | Over capacity | Over capacity | Over capacity |
Requests queued: the server could not run all requests at once, so some waited in a queue. Looped: the model kept repeating the same text, so the benchmark discarded this measurement. 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.| Phase | Context | Requests at once | Status | Total tok/s | Per request tok/s | TTFT | Time between tokens | Average active requests | Completed requests | Output tokens | Measured for |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Prefill | 8K | 1 | Valid | 8,122.0 | — | 1,010 ms | — | — | 1 | — | — |
| Prefill | 16K | 1 | Valid | 6,493.0 | — | 2,504 ms | — | — | 1 | — | — |
| Prefill | 32K | 1 | Valid | 4,448.0 | — | 7,277 ms | — | — | 1 | — | — |
| Prefill | 64K | 1 | Valid | 2,738.0 | — | 23,582 ms | — | — | 1 | — | — |
| Prefill | 128K | 1 | Valid | 1,577.0 | — | 81,815 ms | — | — | 1 | — | — |
| Prefill | 251K | 1 | Valid | 869.0 | — | 291,002 ms | — | — | 1 | — | — |
| Prefill | 256K | 1 | Valid | 853.0 | — | 302,427 ms | — | — | 1 | — | — |
| Sustained decode | 8K | 1 | Valid | 53.3 | 53.4 | 53 ms | 18.7 ms | 1.0 | 0 | 1,599 | 30.0 s |
| Sustained decode | 8K | 2 | Looped | — | — | — | — | — | 0 | 0 | — |
| Sustained decode | 8K | 4 | Looped | — | — | — | — | — | 0 | 0 | — |
| Sustained decode | 8K | 8 | Looped | — | — | — | — | — | 0 | 0 | — |
| Sustained decode | 8K | 16 | Looped | — | — | — | — | — | 0 | 0 | — |
| Sustained decode | 8K | 32 | Valid | 858.3 | 27.3 | 322 ms | 36.6 ms | 32.0 | 0 | 25,827 | 30.1 s |
| Sustained decode | 8K | 64 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 8K | 128 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 16K | 1 | Valid | 50.1 | 50.1 | 76 ms | 19.9 ms | 1.0 | 0 | 1,503 | 30.0 s |
| Sustained decode | 16K | 2 | Valid | 100.2 | 50.1 | 107 ms | 19.9 ms | 2.0 | 0 | 3,005 | 30.0 s |
| Sustained decode | 16K | 4 | Looped | — | — | — | — | — | 0 | 0 | — |
| Sustained decode | 16K | 8 | Looped | — | — | — | — | — | 0 | 0 | — |
| Sustained decode | 16K | 16 | Valid | 442.0 | 27.8 | 324 ms | 35.9 ms | 16.0 | 0 | 13,259 | 30.0 s |
| Sustained decode | 16K | 32 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 16K | 64 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 16K | 128 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 32K | 1 | Valid | 46.3 | 46.4 | 121 ms | 21.6 ms | 1.0 | 0 | 1,389 | 30.0 s |
| Sustained decode | 32K | 2 | Looped | — | — | — | — | — | 0 | 0 | — |
| Sustained decode | 32K | 4 | Looped | — | — | — | — | — | 0 | 0 | — |
| Sustained decode | 32K | 8 | Looped | — | — | — | — | — | 0 | 0 | — |
| Sustained decode | 32K | 16 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 32K | 32 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 32K | 64 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 32K | 128 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 64K | 1 | Valid | 41.9 | 42.0 | 217 ms | 23.8 ms | 1.0 | 0 | 1,257 | 30.0 s |
| Sustained decode | 64K | 2 | Looped | — | — | — | — | — | 0 | 0 | — |
| Sustained decode | 64K | 4 | Looped | — | — | — | — | — | 0 | 0 | — |
| Sustained decode | 64K | 8 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 64K | 16 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 64K | 32 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 64K | 64 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 64K | 128 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 128K | 1 | Valid | 35.2 | 35.3 | 490 ms | 28.3 ms | 1.0 | 0 | 1,056 | 30.0 s |
| Sustained decode | 128K | 2 | Valid | 70.1 | 35.1 | 642 ms | 28.5 ms | 2.0 | 0 | 2,104 | 30.0 s |
| Sustained decode | 128K | 4 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 128K | 8 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 128K | 16 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 128K | 32 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 128K | 64 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 128K | 128 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 251K | 1 | Requests queued | 26.5 | 27.1 | 1,133 ms | 36.9 ms | 1.0 | 0 | 796 | 30.0 s |
| Sustained decode | 251K | 2 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 251K | 4 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 251K | 8 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 251K | 16 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 251K | 32 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 251K | 64 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 251K | 128 | Over capacity | — | — | — | — | — | — | — | — |
Requests queued: the server could not run all requests at once, so some waited in a queue. Looped: the model kept repeating the same text, so the benchmark discarded this measurement. Over capacity: not run because this many requests at this context would exceed the server's KV cache.