Speed test matrix

Qwen3.5-9B-AWQ-4bit on RTX 3090 24GB

vllm · 0.32.0 · W4A16 g326 contexts (0–248K) × 8 concurrency levels (1–128) · prefill at 8K, 16K, 32K, 64K, 128K

by @Sandbergllm-inference-bench v0.8.0Measured Nov 1, 2026, 16:33 UTC

Hardware

Type: Discrete GPU
GPU: RTX 3090
VRAM: 24 GB
CPU: Intel Core i7-9700K (8C)
RAM: 32 GB
OS: Ubuntu 24.04.5

Engine

Engine: vllm
Version: 0.32.0
Quantization: W4A16 g32
Backend: cuda
Repository: github.com/vllm-project/vllm
Build: vllm/vllm-openai:v0.32.0-x86_64

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: MTP, 2 tokens lookahead
# 1x RTX 3090 24GB (Zagitta/gpu-unlock), 420 W limit, PCIe Gen3 x16 | Intel Core i7-9700K 8C | 32 GB DDR4 | Ubuntu 24.04.5, NVIDIA driver 580.178.04
vllm serve cyankiwi/Qwen3.5-9B-AWQ-4bit --max-model-len 262144 --gpu-memory-utilization 0.94 --max-num-seqs 128 --enable-prefix-caching --trust-remote-code --language-model-only --speculative-config {"method":"mtp","num_speculative_tokens":2}

Prefill

prompt tok/s by context length

8K context6,6921,224 ms TTFT

16K context6,3492,555 ms TTFT

32K context5,8555,515 ms TTFT

64K context5,12812,562 ms TTFT

128K context4,04131,847 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	135.340 ms TTFT	257.344 ms TTFT	512.588 ms TTFT	894.2134 ms TTFT	1,413.9251 ms TTFT	1,788.6348 ms TTFT	Over capacity	Over capacity
16K	101.6235 ms TTFT	191.4339 ms TTFT	369.1592 ms TTFT	552.8986 ms TTFT	Over capacity	Over capacity	Over capacity	Over capacity
32K	76.3233 ms TTFT	150.1357 ms TTFT	287.3592 ms TTFT	373.8971 ms TTFT	Over capacity	Over capacity	Over capacity	Over capacity
64K	53.7351 ms TTFT	100.5491 ms TTFT	194.1753 ms TTFT	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity
128K	31.9578 ms TTFT	64.0812 ms TTFT	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity
248K	19.6946 ms TTFT	Over capacity	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	6,692.0	—	1,224 ms	—	—	1	—	—
Prefill	16K	1	Valid	6,349.0	—	2,555 ms	—	—	1	—	—
Prefill	32K	1	Valid	5,855.0	—	5,515 ms	—	—	1	—	—
Prefill	64K	1	Valid	5,128.0	—	12,562 ms	—	—	1	—	—
Prefill	128K	1	Valid	4,041.0	—	31,847 ms	—	—	1	—	—
Sustained decode	0	1	Valid	135.3	136.1	40 ms	7.3 ms	1.0	0	4,060	30.0 s
Sustained decode	0	2	Valid	257.3	132.3	44 ms	7.6 ms	2.0	0	7,717	30.0 s
Sustained decode	0	4	Valid	512.5	131.2	88 ms	7.6 ms	4.0	0	15,373	30.0 s
Sustained decode	0	8	Valid	894.2	114.5	134 ms	8.7 ms	8.0	0	26,826	30.0 s
Sustained decode	0	16	Valid	1,413.9	91.3	251 ms	10.9 ms	16.0	0	42,411	30.0 s
Sustained decode	0	32	Valid	1,788.6	57.4	348 ms	17.4 ms	32.0	0	53,663	30.0 s
Sustained decode	0	64	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	0	128	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	16K	1	Valid	101.6	102.0	235 ms	9.8 ms	1.0	0	3,051	30.0 s
Sustained decode	16K	2	Valid	191.4	96.7	339 ms	10.3 ms	2.0	0	5,743	30.0 s
Sustained decode	16K	4	Valid	369.1	92.6	592 ms	10.8 ms	4.0	0	11,068	30.0 s
Sustained decode	16K	8	Valid	552.8	69.3	986 ms	14.4 ms	8.0	0	16,570	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	76.3	77.9	233 ms	12.8 ms	1.0	0	2,289	30.0 s
Sustained decode	32K	2	Valid	150.1	75.9	357 ms	13.2 ms	2.0	0	4,503	30.0 s
Sustained decode	32K	4	Valid	287.3	74.5	592 ms	13.4 ms	4.0	0	8,611	30.0 s
Sustained decode	32K	8	Valid	373.8	47.2	971 ms	21.2 ms	8.0	0	11,209	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	53.7	54.7	351 ms	18.3 ms	1.0	0	1,610	30.0 s
Sustained decode	64K	2	Valid	100.5	50.9	491 ms	19.7 ms	2.0	0	3,014	30.0 s
Sustained decode	64K	4	Valid	194.1	48.5	753 ms	20.6 ms	4.0	0	5,820	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	31.9	32.8	578 ms	30.4 ms	1.0	0	956	30.0 s
Sustained decode	128K	2	Valid	64.0	32.9	812 ms	30.4 ms	2.0	0	1,920	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	Valid	19.6	20.2	946 ms	49.6 ms	1.0	0	588	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	—	—	—	—	—	—	—	—

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