Speed test matrix

gemma-4-26B-A4B-it-NVFP4 on RTX 3090 24GB

vllm · 0.32.0 · NVFP46 contexts (0–208K) × 8 concurrency levels (1–128) · prefill at 8K, 16K, 32K, 64K, 128K

by @Sandbergllm-inference-bench v0.8.0Measured Oct 30, 2026, 09:05 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: NVFP4
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
# 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 RedHatAI/gemma-4-26B-A4B-it-NVFP4 --max-model-len 221184 --gpu-memory-utilization 0.94 --max-num-seqs 128 --enable-prefix-caching --trust-remote-code --language-model-only

Prefill

prompt tok/s by context length

8K context7,0331,140 ms TTFT

16K context5,8662,707 ms TTFT

32K context4,4407,117 ms TTFT

64K context2,96021,294 ms TTFT

128K context1,78270,673 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	150.230 ms TTFT	256.936 ms TTFT	406.434 ms TTFT	628.841 ms TTFT	929.760 ms TTFT	Over capacity	Over capacity	Over capacity
16K	123.257 ms TTFT	Looped	Looped	Looped	Over capacity	Over capacity	Over capacity	Over capacity
32K	Looped	Looped	Looped	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity
64K	Looped	Looped	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity
128K	80.1400 ms TTFT	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity
208K	64.6781 ms TTFT	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity

Looped: the model kept repeating the same text, so the benchmark discarded this measurement. 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	7,033.0	—	1,140 ms	—	—	1	—	—
Prefill	16K	1	Valid	5,866.0	—	2,707 ms	—	—	1	—	—
Prefill	32K	1	Valid	4,440.0	—	7,117 ms	—	—	1	—	—
Prefill	64K	1	Valid	2,960.0	—	21,294 ms	—	—	1	—	—
Prefill	128K	1	Valid	1,782.0	—	70,673 ms	—	—	1	—	—
Sustained decode	0	1	Valid	150.2	151.6	30 ms	6.6 ms	1.0	0	4,505	30.0 s
Sustained decode	0	2	Valid	256.9	130.0	36 ms	7.7 ms	2.0	0	7,706	30.0 s
Sustained decode	0	4	Valid	406.4	103.9	34 ms	9.6 ms	4.0	0	12,191	30.0 s
Sustained decode	0	8	Valid	628.8	81.0	41 ms	12.3 ms	8.0	0	18,864	30.0 s
Sustained decode	0	16	Valid	929.7	60.5	60 ms	16.5 ms	16.0	0	27,888	30.0 s
Sustained decode	0	32	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	0	64	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	0	128	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	16K	1	Valid	123.2	123.8	57 ms	8.1 ms	1.0	0	3,697	30.0 s
Sustained decode	16K	2	Looped	—	—	—	—	—	0	0	—
Sustained decode	16K	4	Looped	—	—	—	—	—	0	0	—
Sustained decode	16K	8	Looped	—	—	—	—	—	0	0	—
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	Looped	—	—	—	—	—	0	0	—
Sustained decode	32K	2	Looped	—	—	—	—	—	0	0	—
Sustained decode	32K	4	Looped	—	—	—	—	—	0	0	—
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	Looped	—	—	—	—	—	0	0	—
Sustained decode	64K	2	Looped	—	—	—	—	—	0	0	—
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	128K	1	Valid	80.1	80.3	400 ms	12.5 ms	1.0	0	2,403	30.0 s
Sustained decode	128K	2	Over capacity	—	—	—	—	—	—	—	—
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	208K	1	Valid	64.6	64.9	781 ms	15.4 ms	1.0	0	1,938	30.0 s
Sustained decode	208K	2	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	208K	4	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	208K	8	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	208K	16	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	208K	32	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	208K	64	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	208K	128	Over capacity	—	—	—	—	—	—	—	—

Looped: the model kept repeating the same text, so the benchmark discarded this measurement. Over capacity: not run because this many requests at this context would exceed the server's KV cache.