Speed test matrix

Qwen3.5-27B-AWQ-4bit on RTX 3090 24GB

vllm · 0.32.0 · AWQ 4-bit4 contexts (0–56K) × 8 concurrency levels (1–128) · prefill at 8K, 16K, 32K, 56K

by @Sandbergllm-inference-bench v0.8.0Measured Oct 31, 2026, 04:35 UTC

Hardware

Type: Discrete GPU
GPU: RTX 3090
VRAM: 24 GB
CPU: AMD Ryzen 9 7900X (12C)
RAM: 32 GB
OS: Ubuntu 24.04.5

Engine

Engine: vllm
Version: 0.32.0
Quantization: AWQ 4-bit
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: none. KV cache: int4_per_token_head
# 1x RTX 3090 24GB (Zagitta/gpu-unlock), 350 W limit (VBIOS max), PCIe Gen4 x16 | AMD Ryzen 9 7900X 12C | 32 GB DDR5 | Ubuntu 24.04.5, NVIDIA driver 595.99.02
vllm serve cyankiwi/Qwen3.5-27B-AWQ-4bit --max-model-len 65536 --gpu-memory-utilization 0.94 --max-num-seqs 29 --enable-prefix-caching --trust-remote-code --language-model-only --kv-cache-dtype int4_per_token_head --attention-backend TRITON_ATTN --max-cudagraph-capture-size 16

Prefill

prompt tok/s by context length

8K context1,5555,265 ms TTFT

16K context1,30812,402 ms TTFT

32K context1,00332,192 ms TTFT

56K context743.075,937 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	43.272 ms TTFT	80.3133 ms TTFT	155.0217 ms TTFT	173.3Requests queued361 ms TTFT	Over capacity	Over capacity	Over capacity	Over capacity
16K	38.31,010 ms TTFT	68.61,533 ms TTFT	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity
32K	34.62,784 ms TTFT	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity
56K	27.1Requests queued3,613 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	1,555.0	—	5,265 ms	—	—	1	—	—
Prefill	16K	1	Valid	1,308.0	—	12,402 ms	—	—	1	—	—
Prefill	32K	1	Valid	1,003.0	—	32,192 ms	—	—	1	—	—
Prefill	56K	1	Valid	743.0	—	75,937 ms	—	—	1	—	—
Sustained decode	0	1	Valid	43.2	43.2	72 ms	23.1 ms	1.0	0	1,295	30.0 s
Sustained decode	0	2	Valid	80.3	40.2	133 ms	24.9 ms	2.0	0	2,408	30.0 s
Sustained decode	0	4	Valid	155.0	38.8	217 ms	25.8 ms	4.0	0	4,648	30.0 s
Sustained decode	0	8	Requests queued	173.3	33.8	361 ms	29.6 ms	6.3	0	5,197	30.0 s
Sustained decode	0	16	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	0	32	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	0	64	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	0	128	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	16K	1	Valid	38.3	38.3	1,010 ms	26.1 ms	1.0	0	1,150	30.0 s
Sustained decode	16K	2	Valid	68.6	33.9	1,533 ms	29.5 ms	2.0	0	2,058	30.0 s
Sustained decode	16K	4	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	16K	8	Over capacity	—	—	—	—	—	—	—	—
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	34.6	34.6	2,784 ms	28.9 ms	1.0	0	1,037	30.0 s
Sustained decode	32K	2	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	32K	4	Over capacity	—	—	—	—	—	—	—	—
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	56K	1	Requests queued	27.1	30.2	3,613 ms	33.1 ms	0.9	0	813	30.0 s
Sustained decode	56K	2	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	56K	4	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	56K	8	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	56K	16	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	56K	32	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	56K	64	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	56K	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.