Speed test matrix

granite-4.2-30b on H200 NVL 141GB

vllm · 0.25.0 · BF165 contexts (8K–123K) × 8 concurrency levels (1–128) · prefill at 8K, 16K, 32K, 64K, 123K, 128K

by @keennaylmx dev importMeasured Sep 3, 2026, 08:41 UTC

Hardware

Type: Discrete GPU
GPU: H200 NVL
VRAM: 141 GB
OS: linux

Engine

Engine: vllm
Version: 0.25.0
Quantization: BF16
Backend: cuda
Repository: github.com/vllm-project/vllm
Build: cuda-13.2

Run

Served model: granite-4.2-30b
Revision: main
Benchmark tool: llm-inference-bench v0.4.29
Decode test: Fixed duration
Prefill mode: Measured during decode warm-up
Time per measurement: 30 s
Max output: 4,096 tokens

Launch command

"$HOME/env_h200-ibm-vllm/bin/vllm" serve ibm-granite/granite-4.2-30b --served-model-name granite-4.2-30b --tensor-parallel-size 1 --reasoning-parser granite_thinking_parser --reasoning-parser-plugin /workspace/compute/recipes/nvidia-h200/ibm-granite/plugin/granite_thinking_parser.py --enable-auto-tool-choice --tool-call-parser qwen3_coder --max-model-len 131072 --gpu-memory-utilization 0.869345 --host 0.0.0.0 --port 8000 --api-key <redacted> --dtype bfloat16

Prefill

prompt tok/s by context length

8K context9,269884 ms TTFT

16K context8,4691,914 ms TTFT

32K context7,5264,288 ms TTFT

64K context6,12110,520 ms TTFT

123K context4,51427,377 ms TTFT

128K context4,42629,062 ms TTFT

Sustained decode

total tok/s · rows are context length, columns are simultaneous requests

Context	1	2	4	8	16	32	64	128
8K	55.156 ms TTFT	107.686 ms TTFT	208.8117 ms TTFT	387.3166 ms TTFT	672.0265 ms TTFT	Over capacity	Over capacity	Over capacity
16K	53.589 ms TTFT	103.6111 ms TTFT	195.7165 ms TTFT	343.9226 ms TTFT	Over capacity	Over capacity	Over capacity	Over capacity
32K	50.6110 ms TTFT	96.1140 ms TTFT	173.1242 ms TTFT	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity
64K	45.7184 ms TTFT	84.5240 ms TTFT	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity
123K	38.9301 ms TTFT	68.4415 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	9,269.0	—	884 ms	—	—	1	—	—
Prefill	16K	1	Valid	8,469.0	—	1,914 ms	—	—	1	—	—
Prefill	32K	1	Valid	7,526.0	—	4,288 ms	—	—	1	—	—
Prefill	64K	1	Valid	6,121.0	—	10,520 ms	—	—	1	—	—
Prefill	123K	1	Valid	4,514.0	—	27,377 ms	—	—	1	—	—
Prefill	128K	1	Valid	4,426.0	—	29,062 ms	—	—	1	—	—
Sustained decode	8K	1	Valid	55.1	55.1	56 ms	18.1 ms	1.0	0	1,652	30.0 s
Sustained decode	8K	2	Valid	107.6	53.9	86 ms	18.6 ms	2.0	0	3,229	30.0 s
Sustained decode	8K	4	Valid	208.8	52.3	117 ms	19.1 ms	4.0	0	6,264	30.0 s
Sustained decode	8K	8	Valid	387.3	48.5	166 ms	20.6 ms	8.0	0	11,616	30.0 s
Sustained decode	8K	16	Valid	672.0	42.2	265 ms	23.7 ms	16.0	0	20,160	30.0 s
Sustained decode	8K	32	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	8K	64	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	8K	128	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	16K	1	Valid	53.5	53.5	89 ms	18.7 ms	1.0	0	1,604	30.0 s
Sustained decode	16K	2	Valid	103.6	51.9	111 ms	19.3 ms	2.0	0	3,108	30.0 s
Sustained decode	16K	4	Valid	195.7	49.0	165 ms	20.4 ms	4.0	0	5,868	30.0 s
Sustained decode	16K	8	Valid	343.9	43.1	226 ms	23.2 ms	8.0	0	10,312	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	50.6	50.6	110 ms	19.8 ms	1.0	0	1,517	30.0 s
Sustained decode	32K	2	Valid	96.1	48.1	140 ms	20.8 ms	2.0	0	2,884	30.0 s
Sustained decode	32K	4	Valid	173.1	43.3	242 ms	23.1 ms	4.0	0	5,192	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	45.7	45.8	184 ms	21.8 ms	1.0	0	1,371	30.0 s
Sustained decode	64K	2	Valid	84.5	42.3	240 ms	23.7 ms	2.0	0	2,534	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	123K	1	Valid	38.9	39.0	301 ms	25.6 ms	1.0	0	1,168	30.0 s
Sustained decode	123K	2	Valid	68.4	34.3	415 ms	29.2 ms	2.0	0	2,052	30.0 s
Sustained decode	123K	4	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	123K	8	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	123K	16	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	123K	32	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	123K	64	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	123K	128	Over capacity	—	—	—	—	—	—	—	—

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