Speed test matrix

Qwen3-Coder-30B-A3B-Instruct-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

by @Sandbergllm-inference-bench v0.8.0Measured Oct 30, 2026, 04:57 UTC

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-Coder-30B-A3B-Instruct-FP8 --max-model-len 262144 --gpu-memory-utilization 0.94 --max-num-seqs 128 --enable-prefix-caching --trust-remote-code

Prefill

prompt tok/s by context length

8K context8,0961,014 ms TTFT

16K context6,5622,475 ms TTFT

32K context4,8136,719 ms TTFT

64K context3,08420,924 ms TTFT

128K context1,78872,099 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	145.076 ms TTFT	244.4101 ms TTFT	401.796 ms TTFT	646.1134 ms TTFT	1,011.9171 ms TTFT	1,564.2333 ms TTFT	Over capacity	Over capacity
16K	121.9171 ms TTFT	192.1192 ms TTFT	282.7288 ms TTFT	394.6446 ms TTFT	Over capacity	Over capacity	Over capacity	Over capacity
32K	105.1259 ms TTFT	156.3350 ms TTFT	214.9578 ms TTFT	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity
64K	82.2505 ms TTFT	111.7650 ms TTFT	141.2976 ms TTFT	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity
128K	56.9992 ms TTFT	68.91,280 ms TTFT	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity
248K	34.4Requests queued1,870 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	8,096.0	—	1,014 ms	—	—	1	—	—
Prefill	16K	1	Valid	6,562.0	—	2,475 ms	—	—	1	—	—
Prefill	32K	1	Valid	4,813.0	—	6,719 ms	—	—	1	—	—
Prefill	64K	1	Valid	3,084.0	—	20,924 ms	—	—	1	—	—
Prefill	128K	1	Valid	1,788.0	—	72,099 ms	—	—	1	—	—
Sustained decode	0	1	Valid	145.0	146.0	76 ms	6.9 ms	1.0	0	4,350	30.0 s
Sustained decode	0	2	Valid	244.4	123.5	101 ms	8.1 ms	2.0	0	7,332	30.0 s
Sustained decode	0	4	Valid	401.7	101.9	96 ms	9.8 ms	4.0	0	12,050	30.0 s
Sustained decode	0	8	Valid	646.1	82.5	134 ms	12.1 ms	8.0	0	19,384	30.0 s
Sustained decode	0	16	Valid	1,011.9	65.2	171 ms	15.3 ms	16.0	0	30,352	30.0 s
Sustained decode	0	32	Valid	1,564.2	50.9	333 ms	19.7 ms	32.0	0	46,912	30.0 s
Sustained decode	0	64	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	0	128	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	16K	1	Valid	121.9	122.3	171 ms	8.2 ms	1.0	0	3,658	30.0 s
Sustained decode	16K	2	Valid	192.1	96.5	192 ms	10.4 ms	2.0	0	5,766	30.0 s
Sustained decode	16K	4	Valid	282.7	71.2	288 ms	14.0 ms	4.0	0	8,481	30.0 s
Sustained decode	16K	8	Valid	394.6	49.5	446 ms	20.2 ms	8.0	0	11,832	30.0 s
Sustained decode	16K	16	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	16K	32	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	16K	64	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	16K	128	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	32K	1	Valid	105.1	105.6	259 ms	9.5 ms	1.0	0	3,153	30.0 s
Sustained decode	32K	2	Valid	156.3	78.5	350 ms	12.7 ms	2.0	0	4,688	30.0 s
Sustained decode	32K	4	Valid	214.9	54.0	578 ms	18.5 ms	4.0	0	6,444	30.0 s
Sustained decode	32K	8	Over capacity	—	—	—	—	—	—	—	—
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	82.2	82.5	505 ms	12.1 ms	1.0	0	2,464	30.0 s
Sustained decode	64K	2	Valid	111.7	56.1	650 ms	17.8 ms	2.0	0	3,350	30.0 s
Sustained decode	64K	4	Valid	141.2	35.3	976 ms	28.3 ms	4.0	0	4,236	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	56.9	57.2	992 ms	17.5 ms	1.0	0	1,707	30.0 s
Sustained decode	128K	2	Valid	68.9	34.6	1,280 ms	28.9 ms	2.0	0	2,066	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	34.4	36.0	1,870 ms	27.7 ms	0.9	0	1,032	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.