Speed test matrix
granite-4.2-3b-nvfp4 on RTX 3090 24GB
vllm · 0.32.0 · NVFP45 contexts (0–120K) × 8 concurrency levels (1–128) · prefill at 8K, 16K, 32K, 64K, 120K
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
- 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), 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 ibm-granite/granite-4.2-3b-nvfp4 --max-model-len 131072 --gpu-memory-utilization 0.94 --max-num-seqs 128 --enable-prefix-caching --trust-remote-code
Prefill
prompt tok/s by context length8K context12.1K680 ms TTFT
16K context9,7611,671 ms TTFT
32K context7,2864,457 ms TTFT
64K context4,83113,407 ms TTFT
120K context3,01340,264 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 | 206.319 ms TTFT | 374.422 ms TTFT | 667.919 ms TTFT | 1,114.823 ms TTFT | 1,775.852 ms TTFT | Over capacity | Over capacity | Over capacity |
| 16K | 157.968 ms TTFT | 254.473 ms TTFT | 426.9110 ms TTFT | 535.3181 ms TTFT | Over capacity | Over capacity | Over capacity | Over capacity |
| 32K | 128.498 ms TTFT | 189.2129 ms TTFT | 300.8198 ms TTFT | Over capacity | Over capacity | Over capacity | Over capacity | Over capacity |
| 64K | 93.1217 ms TTFT | 124.7254 ms TTFT | Over capacity | Over capacity | Over capacity | Over capacity | Over capacity | Over capacity |
| 120K | 62.7402 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 | 12.1K | — | 680 ms | — | — | 1 | — | — |
| Prefill | 16K | 1 | Valid | 9,761.0 | — | 1,671 ms | — | — | 1 | — | — |
| Prefill | 32K | 1 | Valid | 7,286.0 | — | 4,457 ms | — | — | 1 | — | — |
| Prefill | 64K | 1 | Valid | 4,831.0 | — | 13,407 ms | — | — | 1 | — | — |
| Prefill | 120K | 1 | Valid | 3,013.0 | — | 40,264 ms | — | — | 1 | — | — |
| Sustained decode | 0 | 1 | Valid | 206.3 | 209.7 | 19 ms | 4.8 ms | 1.0 | 0 | 6,190 | 30.0 s |
| Sustained decode | 0 | 2 | Valid | 374.4 | 192.1 | 22 ms | 5.2 ms | 2.0 | 0 | 11,230 | 30.0 s |
| Sustained decode | 0 | 4 | Valid | 667.9 | 173.3 | 19 ms | 5.8 ms | 4.0 | 0 | 20,036 | 30.0 s |
| Sustained decode | 0 | 8 | Valid | 1,114.8 | 147.0 | 23 ms | 6.8 ms | 8.0 | 0 | 33,440 | 30.0 s |
| Sustained decode | 0 | 16 | Valid | 1,775.8 | 119.7 | 52 ms | 8.4 ms | 16.0 | 0 | 53,279 | 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 | 157.9 | 158.7 | 68 ms | 6.3 ms | 1.0 | 0 | 4,737 | 30.0 s |
| Sustained decode | 16K | 2 | Valid | 254.4 | 128.0 | 73 ms | 7.8 ms | 2.0 | 0 | 7,632 | 30.0 s |
| Sustained decode | 16K | 4 | Valid | 426.9 | 107.8 | 110 ms | 9.3 ms | 4.0 | 0 | 12,808 | 30.0 s |
| Sustained decode | 16K | 8 | Valid | 535.3 | 68.1 | 181 ms | 14.7 ms | 8.0 | 0 | 16,056 | 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 | 128.4 | 128.8 | 98 ms | 7.8 ms | 1.0 | 0 | 3,853 | 30.0 s |
| Sustained decode | 32K | 2 | Valid | 189.2 | 94.9 | 129 ms | 10.5 ms | 2.0 | 0 | 5,680 | 30.0 s |
| Sustained decode | 32K | 4 | Valid | 300.8 | 75.3 | 198 ms | 13.3 ms | 4.0 | 0 | 9,020 | 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 | 93.1 | 93.4 | 217 ms | 10.7 ms | 1.0 | 0 | 2,793 | 30.0 s |
| Sustained decode | 64K | 2 | Valid | 124.7 | 62.5 | 254 ms | 16.0 ms | 2.0 | 0 | 3,742 | 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 | 120K | 1 | Valid | 62.7 | 62.9 | 402 ms | 15.9 ms | 1.0 | 0 | 1,880 | 30.0 s |
| Sustained decode | 120K | 2 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 120K | 4 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 120K | 8 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 120K | 16 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 120K | 32 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 120K | 64 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 120K | 128 | Over capacity | — | — | — | — | — | — | — | — |
Over capacity: not run because this many requests at this context would exceed the server's KV cache.