⚡ Performance and Efficiency Benchmarks
This section reports the performance of Qwen 3.5 on NPU with FastFlowLM (FLM).
Note:
- Results are based on FastFlowLM v0.9.36.
- Under FLM’s default NPU power mode (Performance)
- Newer versions may deliver improved performance.
- Fine-tuned models show performance comparable to their base models.
Test System 1:
AMD Ryzen™ AI 7 350 (Kraken Point) with 32 GB DRAM; performance is comparable to other Kraken Point systems.
🚀 Decoding Speed (TPS, or Tokens per Second, starting @ different context lengths)
| Model | HW | 1k | 2k | 4k | 8k | 16k | 32k |
|---|---|---|---|---|---|---|---|
| Qwen 3.5 4B | NPU (FLM) | 15.0 | 14.6 | 14.2 | 13.3 | 11.8 | 9.6 |
🚀 Prefill Speed (TPS, or Tokens per Second, with different prompt lengths)
| Model | HW | 1k | 2k | 4k | 8k | 16k | 32k |
|---|---|---|---|---|---|---|---|
| Qwen 3.5 4B | NPU (FLM) | 378 | 440 | 479 | 493 | 487 | 450 |
🚀 Prefill TTFT with Image Input (Seconds)
Prefill time-to-first-token (TTFT) for Qwen3.5-4B on NPU (FastFlowLM) with different image resolutions.
Mid Resolution Images:
| Model | HW | 720p (1280×720) | 1080p (1920×1080) |
|---|---|---|---|
| Qwen3.5-4B | NPU (FLM) | 3.7 | 7.5 |
High Resolution Images:
| Model | HW | 2K (2560×1440) | 4K (3840×2160) |
|---|---|---|---|
| Qwen3.5-4B | NPU (FLM) | 14.7 | 41.3 |
This test uses a short prompt: “Describe this image.”