Speed test matrix

deepseek-r1-distill-qwen-14b-awq on CMP 170HX 64GB

vllm · 0.32.0rc1.dev249+g18305948a · AWQ5 contexts (0–120K) × 8 concurrency levels (1–128) · prefill at 8K, 16K, 32K, 64K, 120K

by @Sandbergllm-inference-bench v0.8.0Measured Oct 29, 2026, 03:23 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: AWQ
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 casperhansen/deepseek-r1-distill-qwen-14b-awq --max-model-len 131072 --gpu-memory-utilization 0.94 --max-num-seqs 128 --enable-prefix-caching --trust-remote-code

Prefill

prompt tok/s by context length

8K context3,4272,392 ms TTFT

16K context2,9805,447 ms TTFT

32K context2,45113,192 ms TTFT

64K context1,79935,866 ms TTFT

120K context1,22198,989 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	92.666 ms TTFT	175.991 ms TTFT	333.393 ms TTFT	606.4143 ms TTFT	974.8186 ms TTFT	1,487.7259 ms TTFT	Over capacity	Over capacity
16K	73.7151 ms TTFT	124.2193 ms TTFT	194.8314 ms TTFT	271.7424 ms TTFT	Over capacity	Over capacity	Over capacity	Over capacity
32K	61.3276 ms TTFT	94.0366 ms TTFT	131.9574 ms TTFT	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity
64K	46.0452 ms TTFT	62.1696 ms TTFT	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity
120K	31.9953 ms TTFT	38.71,258 ms TTFT	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity	Over 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.

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	3,427.0	—	2,392 ms	—	—	1	—	—
Prefill	16K	1	Valid	2,980.0	—	5,447 ms	—	—	1	—	—
Prefill	32K	1	Valid	2,451.0	—	13,192 ms	—	—	1	—	—
Prefill	64K	1	Valid	1,799.0	—	35,866 ms	—	—	1	—	—
Prefill	120K	1	Valid	1,221.0	—	98,989 ms	—	—	1	—	—
Sustained decode	0	1	Valid	92.6	93.2	66 ms	10.7 ms	1.0	0	2,777	30.0 s
Sustained decode	0	2	Valid	175.9	88.8	91 ms	11.3 ms	2.0	0	5,276	30.0 s
Sustained decode	0	4	Valid	333.3	84.8	93 ms	11.8 ms	4.0	0	9,996	30.0 s
Sustained decode	0	8	Valid	606.4	77.9	143 ms	12.8 ms	8.0	0	18,192	30.0 s
Sustained decode	0	16	Valid	974.8	63.3	186 ms	15.8 ms	16.0	0	29,232	30.0 s
Sustained decode	0	32	Valid	1,487.7	48.7	259 ms	20.5 ms	32.0	0	44,641	30.0 s
Sustained decode	0	64	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	0	128	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	16K	1	Valid	73.7	74.0	151 ms	13.5 ms	1.0	0	2,211	30.0 s
Sustained decode	16K	2	Valid	124.2	62.3	193 ms	16.0 ms	2.0	0	3,726	30.0 s
Sustained decode	16K	4	Valid	194.8	48.9	314 ms	20.5 ms	4.0	0	5,844	30.0 s
Sustained decode	16K	8	Valid	271.7	33.9	424 ms	29.5 ms	8.0	0	8,152	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	61.3	61.4	276 ms	16.3 ms	1.0	0	1,837	30.0 s
Sustained decode	32K	2	Valid	94.0	47.1	366 ms	21.2 ms	2.0	0	2,820	30.0 s
Sustained decode	32K	4	Valid	131.9	33.0	574 ms	30.3 ms	4.0	0	3,956	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	46.0	46.1	452 ms	21.7 ms	1.0	0	1,379	30.0 s
Sustained decode	64K	2	Valid	62.1	31.1	696 ms	32.1 ms	2.0	0	1,860	30.0 s
Sustained decode	64K	4	Over capacity	—	—	—	—	—	—	—	—
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	120K	1	Valid	31.9	32.0	953 ms	31.3 ms	1.0	0	955	30.0 s
Sustained decode	120K	2	Valid	38.7	19.4	1,258 ms	51.5 ms	2.0	0	1,160	30.0 s
Sustained decode	120K	4	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	120K	8	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	120K	16	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	120K	32	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	120K	64	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	120K	128	Over capacity	—	—	—	—	—	—	—	—

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