Home › Benchmarks › Rockchip RK3588 NPU
Benchmark Explorer · Rockchip RK3588 NPU
Rockchip RK3588 NPU: 25 benchmark rows, 0% measured.
Method v1.0 · dataset 2026-09-07 · verified 2026-09-07 · last updated September 2026
0 of 25 rows are measured and published (class C); the rest are EdgeAIStack-derived (class D). 14 model families. Precisions: INT8 (25). Runtimes: rknn (25). Tasks: detection (17), classification (2), pose (3), segmentation (3).
Family × precision: best value
The fastest matching row per family and precision (fps, or tokens/s for language models), with its evidence class. Click a cell for the full breakdown.
| Family | INT8 |
|---|---|
| DeepLabV3 | 34 fps D |
| MobileNet | 350 fps D |
| MobileNetV2 | 405.6 fps D |
| ResNet | 104.6 fps D |
| RetinaFace | 227.2 fps D |
| SSD MobileNet | 125 fps D |
| YOLO11 | 50 fps D |
| YOLO11 | 45.7 fps D |
| YOLO11 | 49.8 fps D |
| YOLOv10 | 50.8 fps D |
| YOLOv5 | 68.5 fps D |
| YOLOv8 | 61 fps D |
| YOLOv8 | 46.4 fps D |
| YOLOX | 35.6 fps D |
Every model on Rockchip RK3588 NPU
DeepLabV3 Mobilenetv2
DeepLabV3 · mobilenetv2
MobileNetv1
MobileNet · v1
MobileNetV2v2
MobileNetV2 · v2
ResNet50
ResNet · 50
RetinaFace Mobile
RetinaFace · mobile
SSD MobileNetv1
SSD MobileNet · v1
YOLO11l
YOLO11 · l
YOLO11m
YOLO11 · m
YOLO11n
YOLO11 · n
YOLO11s
YOLO11 · s
YOLO11x
YOLO11 · x
YOLO11n
YOLO11 · n
YOLO11s
YOLO11 · s
YOLO11m
YOLO11 · m
YOLO11n
YOLO11 · n
YOLO11s
YOLO11 · s
YOLOv10n
YOLOv10 · n
YOLOv5m
YOLOv5 · m
YOLOv5n
YOLOv5 · n
YOLOv5s
YOLOv5 · s
YOLOv8m
YOLOv8 · m
YOLOv8n
YOLOv8 · n
YOLOv8s
YOLOv8 · s
YOLOv8n
YOLOv8 · n
YOLOXs
YOLOX · s
Links
- Live engine: https://edgeaistack.ai/engines/benchmark-explorer/?hw=rk3588_npu&model=deeplabv3&variant=mobilenetv2&mv=1.0&dv=2026-09-07
- Methodology: /methodology/benchmark-explorer/
- Dataset: /datasets/benchmarks-v2.json
- All hardware: /benchmarks/
Query this hardware live.
Filter by model, precision or runtime and get the matching rows with a permanent link.