Speed test matrix

Phi-4-reasoning-plus-NVFP4 on RTX 3090 24GB

vllm · 0.32.0 · NVFP43 contexts (0–24K) × 8 concurrency levels (1–128) · prefill at 8K, 16K, 24K

by @Sandbergllm-inference-bench v0.8.0Measured Oct 30, 2026, 06:36 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 nvidia/Phi-4-reasoning-plus-NVFP4 --max-model-len 32768 --gpu-memory-utilization 0.94 --max-num-seqs 128 --enable-prefix-caching --trust-remote-code

Prefill

prompt tok/s by context length

8K context3,2652,585 ms TTFT

16K context2,9205,656 ms TTFT

24K context2,6389,320 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	80.742 ms TTFT	158.626 ms TTFT	306.234 ms TTFT	574.139 ms TTFT	Over capacity	Over capacity	Over capacity	Over capacity
16K	69.841 ms TTFT	133.560 ms TTFT	225.1101 ms TTFT	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity
24K	65.655 ms TTFT	122.874 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	3,265.0	—	2,585 ms	—	—	1	—	—
Prefill	16K	1	Valid	2,920.0	—	5,656 ms	—	—	1	—	—
Prefill	24K	1	Valid	2,638.0	—	9,320 ms	—	—	1	—	—
Sustained decode	0	1	Valid	80.7	80.9	42 ms	12.4 ms	1.0	0	2,420	30.0 s
Sustained decode	0	2	Valid	158.6	79.7	26 ms	12.5 ms	2.0	0	4,758	30.0 s
Sustained decode	0	4	Valid	306.2	77.2	34 ms	13.0 ms	4.0	0	9,184	30.0 s
Sustained decode	0	8	Valid	574.1	72.8	39 ms	13.7 ms	8.0	0	17,216	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	69.8	69.9	41 ms	14.3 ms	1.0	0	2,094	30.0 s
Sustained decode	16K	2	Valid	133.5	66.9	60 ms	15.0 ms	2.0	0	4,004	30.0 s
Sustained decode	16K	4	Valid	225.1	56.5	101 ms	17.7 ms	4.0	0	6,752	30.0 s
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	24K	1	Valid	65.6	65.6	55 ms	15.2 ms	1.0	0	1,966	30.0 s
Sustained decode	24K	2	Valid	122.8	61.5	74 ms	16.3 ms	2.0	0	3,684	30.0 s
Sustained decode	24K	4	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	24K	8	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	24K	16	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	24K	32	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	24K	64	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	24K	128	Over capacity	—	—	—	—	—	—	—	—

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