Speed test matrix
Qwen3.5-27B-FP8 on CMP 170HX 64GB
vllm · 0.32.0rc1.dev249+g18305948a · FP86 contexts (0–248K) × 8 concurrency levels (1–128) · prefill at 8K, 16K, 32K, 64K, 128K
Hardware
- Type
- Discrete GPU
- GPU
- CMP 170HX
- VRAM
- 64 GB
- CPU
- 2x Intel Xeon E5-2630 v3
- RAM
- 64 GB
- OS
- Ubuntu 26.04.1
Engine
- Engine
- vllm
- Version
- 0.32.0rc1.dev249+g18305948a
- Quantization
- FP8
- Backend
- cuda
- Repository
- github.com/vllm-project/vllm
- Build
- ghcr.io/morrowmake/vllm-cmp170hx@sha256:5b45c4d6feafefe2eba962375b0a9a0b64fd22b9e05abb4d1966f14a32a2db79
Run
- Served model
- bench
- Revision
- main
- Benchmark tool
- llm-inference-bench v0.8.0
- Decode test
- Fixed duration
- Prefill mode
- Measured during decode warm-up
- Time per measurement
- 30 s
- Max output
- 8,192 tokens
Launch command
# Untuned out-of-the-box baseline. Spec decode: none # 1x NVIDIA CMP 170HX 64GB (cmpunlocker), 180 W limit, PCIe Gen2 x4 | 2x Intel Xeon E5-2630 v3 | 64 GB DDR4 | Ubuntu 26.04.1, NVIDIA driver 610.57.04 | vLLM: Morrowmake vllm-cmp170hx image vllm serve Qwen/Qwen3.5-27B-FP8 --max-model-len 262144 --gpu-memory-utilization 0.94 --max-num-seqs 128 --enable-prefix-caching --trust-remote-code --language-model-only
Prefill
prompt tok/s by context length8K context1,7254,721 ms TTFT
16K context1,6319,882 ms TTFT
32K context1,53420,913 ms TTFT
64K context1,38146,371 ms TTFT
128K context1,141112,092 ms TTFT
Sustained decode
total tok/s · rows are context length, columns are simultaneous requests| Context | 1 | 2 | 4 | 8 | 16 | 32 | 64 | 128 |
|---|---|---|---|---|---|---|---|---|
| 0 | 37.4199 ms TTFT | 73.2350 ms TTFT | 141.3426 ms TTFT | 269.6571 ms TTFT | 483.3951 ms TTFT | 796.61,632 ms TTFT | Over capacity | Over capacity |
| 16K | 35.7451 ms TTFT | 68.5601 ms TTFT | 126.6969 ms TTFT | 221.51,651 ms TTFT | 347.43,089 ms TTFT | Over capacity | Over capacity | Over capacity |
| 32K | 34.8754 ms TTFT | 64.61,046 ms TTFT | 114.21,657 ms TTFT | 187.32,879 ms TTFT | Over capacity | Over capacity | Over capacity | Over capacity |
| 64K | 32.71,132 ms TTFT | 58.01,178 ms TTFT | 95.61,756 ms TTFT | Over capacity | Over capacity | Over capacity | Over capacity | Over capacity |
| 128K | 29.31,030 ms TTFT | 47.91,259 ms TTFT | Over capacity | Over capacity | Over capacity | Over capacity | Over capacity | Over capacity |
| 248K | 22.8Requests queued2,683 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. 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 | 1,725.0 | — | 4,721 ms | — | — | 1 | — | — |
| Prefill | 16K | 1 | Valid | 1,631.0 | — | 9,882 ms | — | — | 1 | — | — |
| Prefill | 32K | 1 | Valid | 1,534.0 | — | 20,913 ms | — | — | 1 | — | — |
| Prefill | 64K | 1 | Valid | 1,381.0 | — | 46,371 ms | — | — | 1 | — | — |
| Prefill | 128K | 1 | Valid | 1,141.0 | — | 112,092 ms | — | — | 1 | — | — |
| Sustained decode | 0 | 1 | Valid | 37.4 | 37.4 | 199 ms | 26.7 ms | 1.0 | 0 | 1,122 | 30.0 s |
| Sustained decode | 0 | 2 | Valid | 73.2 | 36.6 | 350 ms | 27.3 ms | 2.0 | 0 | 2,194 | 30.0 s |
| Sustained decode | 0 | 4 | Valid | 141.3 | 35.4 | 426 ms | 28.3 ms | 4.0 | 0 | 4,236 | 30.0 s |
| Sustained decode | 0 | 8 | Valid | 269.6 | 33.8 | 571 ms | 29.6 ms | 8.0 | 0 | 8,080 | 30.0 s |
| Sustained decode | 0 | 16 | Valid | 483.3 | 30.3 | 951 ms | 33.0 ms | 16.0 | 0 | 14,496 | 30.0 s |
| Sustained decode | 0 | 32 | Valid | 796.6 | 24.9 | 1,632 ms | 40.2 ms | 32.0 | 0 | 23,904 | 30.0 s |
| Sustained decode | 0 | 64 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 0 | 128 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 16K | 1 | Valid | 35.7 | 35.7 | 451 ms | 28.0 ms | 1.0 | 0 | 1,070 | 30.0 s |
| Sustained decode | 16K | 2 | Valid | 68.5 | 34.2 | 601 ms | 29.3 ms | 2.0 | 0 | 2,056 | 30.0 s |
| Sustained decode | 16K | 4 | Valid | 126.6 | 31.3 | 969 ms | 32.0 ms | 4.0 | 0 | 3,796 | 30.0 s |
| Sustained decode | 16K | 8 | Valid | 221.5 | 26.9 | 1,651 ms | 37.2 ms | 8.0 | 0 | 6,646 | 30.0 s |
| Sustained decode | 16K | 16 | Valid | 347.4 | 20.3 | 3,089 ms | 49.2 ms | 16.0 | 0 | 10,416 | 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 | 34.8 | 34.9 | 754 ms | 28.7 ms | 1.0 | 0 | 1,044 | 30.0 s |
| Sustained decode | 32K | 2 | Valid | 64.6 | 32.1 | 1,046 ms | 31.1 ms | 2.0 | 0 | 1,938 | 30.0 s |
| Sustained decode | 32K | 4 | Valid | 114.2 | 28.0 | 1,657 ms | 35.8 ms | 4.0 | 0 | 3,424 | 30.0 s |
| Sustained decode | 32K | 8 | Valid | 187.3 | 22.2 | 2,879 ms | 45.0 ms | 8.0 | 0 | 5,618 | 30.0 s |
| 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 | 32.7 | 32.8 | 1,132 ms | 30.5 ms | 1.0 | 0 | 982 | 30.0 s |
| Sustained decode | 64K | 2 | Valid | 58.0 | 28.9 | 1,178 ms | 34.7 ms | 2.0 | 0 | 1,740 | 30.0 s |
| Sustained decode | 64K | 4 | Valid | 95.6 | 23.4 | 1,756 ms | 42.7 ms | 4.0 | 0 | 2,864 | 30.0 s |
| 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 | 29.3 | 29.4 | 1,030 ms | 34.0 ms | 1.0 | 0 | 879 | 30.0 s |
| Sustained decode | 128K | 2 | Valid | 47.9 | 24.0 | 1,259 ms | 41.7 ms | 2.0 | 0 | 1,434 | 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 | 248K | 1 | Requests queued | 22.8 | 24.5 | 2,683 ms | 40.8 ms | 0.9 | 0 | 682 | 30.0 s |
| Sustained decode | 248K | 2 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 248K | 4 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 248K | 8 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 248K | 16 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 248K | 32 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 248K | 64 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 248K | 128 | Over capacity | — | — | — | — | — | — | — | — |
Requests queued: the server could not run all requests at once, so some waited in a queue. Over capacity: not run because this many requests at this context would exceed the server's KV cache.