Qengineering/Qwen3.5-9B-rk3588

Qwen3.5-9B for RK3588 NPU

This repository provides a hardware-accelerated port of Qwen3.5-9B optimized for Rockchip RK3588 NPU.

User:<image>Describe the image.

Answer:The image depicts a humorous and surreal scene set on the surface of the Moon:

• An astronaut in a white spacesuit is reclining casually, as if relaxing or taking a break.
• The astronaut’s helmet reflects the Earth above them — showing a view of our planet from lunar orbit.
• In their gloved hand, they’re holding a green glass beer bottle (possibly resembling a Heineken).
• Their feet are propped up on what appears to be a cooler or crate labeled “Heineken” (though it’s stylized and hard to read clearly).
• Behind them, the Earth looms large in the sky — blue oceans, white clouds, and green landmass visible.
• To the right, part of a lunar ladder or structure can be seen, suggesting this might be near a space base or landing site.
• The ground is rocky and dusty, typical of the lunar surface.

The overall tone is whimsical and satirical — imagining an astronaut enjoying a beer on the Moon, blending science fiction with everyday leisure culture. It’s likely meant to evoke humor rather than realism.

Note: This image is not scientifically accurate — no such beverage or relaxation would be possible in actual space due to lack of atmosphere, extreme temperatures, radiation, and microgravity effects. But as art, it’s fun! 🌕🍺

---

Model Files

Component	File	Precision
LLM	qwen3.5-9b_w8a8_rk3588.rkllm	W8A8
Vision Encoder	qwen3.5-9b-vision_rk3588.rknn	FP16

Hardware Requirements

• Rockchip RK3588 / RK3588S
• RKNPU2 driver
• Tested on:
  • Rock 5C
  • Ubuntu 22.04 / 24.04 (Joshua Riek)

Runtime Requirements

• RKLLM runtime (version 1.3.0)
• RKNN runtime (rknpu2)
• OpenCV (for image preprocessing)

Model performance benchmark (FPS)

All models, with C++ examples, can be found on the Q-engineering GitHub.

All LLM models are quantized to w8a8, while the VLM vision encoders use fp16.

model	RAM (GB)1	llm cold sec2	llm warm sec3	vlm cold sec2	vlm warm sec3	Resolution	Tokens/s
Qwen3.5-9B	9.2	97.1	97.1	11.5	11.5	448 x 448	3.2
Qwen3.5-4B	5.4	52.8	6.2	8.5	0.9	448 x 448	5.2
Qwen3.5-2B	2.9	23.9	3.2	8.5	0.8	448 x 448	11.0
Qwen3.5-0.8B	1.3	10.6	1.9	2.7	0.2	448 x 448	21.6
Qwen3-2B	3.1	21.9	2.6	10.0	0.9	448 x 448	11.5
Qwen3-4B	8.7	49.6	5.6	10.6	1.1	448 x 448	5.7
InternVL3.5-1B	1.9	8.3	8.0	1.5	0.8	448 x 448	24
InternVL3.5-2B	3.0	22	8.0	2.7	0.8	448 x 448	11.2
InternVL3.5-4B	5.4	50	8.0	5.9	0.8	448 x 448	5
InternVL3.5-8B	8.8	92	8.0	50.5	5.8	448 x 448	3.5
Qwen2.5-3B	4.8	48.3	4.0	17.9	1.8	392 x 392	7.0
Qwen2-7B	8.7	86.6	34.5	37.1	20.7	392 x 392	3.7
Qwen2-2.2B	3.3	29.1	2.5	17.1	1.7	392 x 392	12.5
InternVL3-1B	1.3	6.8	1.1	7.8	0.75	448 x 448	30
SmolVLM2-2.2B	3.4	21.2	2.6	10.5	0.9	384 x 384	11
SmolVLM2-500M	0.8	4.8	0.7	2.5	0.25	384 x 384	31
SmolVLM2-256M	0.5	1.1	0.4	2.5	0.25	384 x 384	54

¹ The total used memory; LLM plus the VLM.
² When an llm/vlm model is loaded for the first time from your disk to RAM or NPU, it is called a cold start.
The duration depends on your OS, I/O transfer rate, and memory mapping.
³ Subsequent loading (warm start) takes advantage of the already mapped data in RAM. Mostly, only a few pointers need to be restored.

Example Usage

• see: https://github.com/Qengineering/Qwen3.5-9B-NPU

Notes

• KV Cache up to 16384
• This is not a Transformers-compatible model
• This repository provides precompiled NPU binaries
• CPU fallback is not supported