Speed test matrix
granite-4.2-30b-nvfp4 on RTX 3090 24GB
vllm · 0.32.0 · NVFP44 contexts (0–48K) × 8 concurrency levels (1–128) · prefill at 8K, 16K, 32K, 48K
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. 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 ibm-granite/granite-4.2-30b-nvfp4 --max-model-len 57344 --gpu-memory-utilization 0.94 --max-num-seqs 128 --enable-prefix-caching --trust-remote-code --kv-cache-dtype int4_per_token_head --attention-backend TRITON_ATTN
Prefill
prompt tok/s by context length8K context1,1737,025 ms TTFT
16K context923.017,682 ms TTFT
32K context657.049,416 ms TTFT
48K context508.095,664 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 | 39.380 ms TTFT | 77.593 ms TTFT | 148.259 ms TTFT | 278.463 ms TTFT | Over capacity | Over capacity | Over capacity | Over capacity |
| 16K | 28.5130 ms TTFT | 52.0211 ms TTFT | Over capacity | Over capacity | Over capacity | Over capacity | Over capacity | Over capacity |
| 32K | 22.4225 ms TTFT | Over capacity | Over capacity | Over capacity | Over capacity | Over capacity | Over capacity | Over capacity |
| 48K | 18.5Requests queued330 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 | 1,173.0 | — | 7,025 ms | — | — | 1 | — | — |
| Prefill | 16K | 1 | Valid | 923.0 | — | 17,682 ms | — | — | 1 | — | — |
| Prefill | 32K | 1 | Valid | 657.0 | — | 49,416 ms | — | — | 1 | — | — |
| Prefill | 48K | 1 | Valid | 508.0 | — | 95,664 ms | — | — | 1 | — | — |
| Sustained decode | 0 | 1 | Valid | 39.3 | 39.5 | 80 ms | 25.3 ms | 1.0 | 0 | 1,180 | 30.0 s |
| Sustained decode | 0 | 2 | Valid | 77.5 | 38.9 | 93 ms | 25.7 ms | 2.0 | 0 | 2,322 | 30.0 s |
| Sustained decode | 0 | 4 | Valid | 148.2 | 37.3 | 59 ms | 26.8 ms | 4.0 | 0 | 4,444 | 30.0 s |
| Sustained decode | 0 | 8 | Valid | 278.4 | 35.2 | 63 ms | 28.4 ms | 8.0 | 0 | 8,352 | 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 | 28.5 | 28.5 | 130 ms | 35.1 ms | 1.0 | 0 | 854 | 30.0 s |
| Sustained decode | 16K | 2 | Valid | 52.0 | 26.0 | 211 ms | 38.5 ms | 2.0 | 0 | 1,558 | 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 | 22.4 | 22.4 | 225 ms | 44.6 ms | 1.0 | 0 | 672 | 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 | 48K | 1 | Requests queued | 18.5 | 18.5 | 330 ms | 54.1 ms | 1.0 | 0 | 554 | 30.0 s |
| Sustained decode | 48K | 2 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 48K | 4 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 48K | 8 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 48K | 16 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 48K | 32 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 48K | 64 | Over capacity | — | — | — | — | — | — | — | — |
| Sustained decode | 48K | 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.