ConvNext-Base: Optimized for Qualcomm Devices

ConvNextBase is a machine learning model that can classify images from the Imagenet dataset. It can also be used as a backbone in building more complex models for specific use cases.

This is based on the implementation of ConvNext-Base found here. This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the Qualcomm® AI Hub Models library to export with custom configurations. More details on model performance across various devices, can be found here.

Qualcomm AI Hub Models uses Qualcomm AI Hub Workbench to compile, profile, and evaluate this model. Sign up to run these models on a hosted Qualcomm® device.

Getting Started

There are two ways to deploy this model on your device:

Option 1: Download Pre-Exported Models

Below are pre-exported model assets ready for deployment.

Runtime	Precision	Chipset	SDK Versions	Download
ONNX	float	Universal	QAIRT 2.45, ONNX Runtime 1.27.1	Download
ONNX	w8a16	Universal	QAIRT 2.45, ONNX Runtime 1.27.1	Download
QNN_DLC	float	Universal	QAIRT 2.45	Download
QNN_DLC	w8a16	Universal	QAIRT 2.45	Download
TFLITE	float	Universal	QAIRT 2.45	Download

For more device-specific assets and performance metrics, visit ConvNext-Base on Qualcomm® AI Hub.

Option 2: Export with Custom Configurations

Use the Qualcomm® AI Hub Models Python library to compile and export the model with your own:

• Custom weights (e.g., fine-tuned checkpoints)
• Custom input shapes
• Target device and runtime configurations

This option is ideal if you need to customize the model beyond the default configuration provided here.

See our repository for ConvNext-Base on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.image_classification

Model Stats:

• Input resolution: 224x224
• Model checkpoint: Imagenet
• Model size (float): 338 MB
• Model size (w8a16): 88.7 MB
• Number of parameters: 88.6M

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
ConvNext-Base	ONNX	float	Snapdragon® X2 Elite	3.513 ms	2 - 2 MB	NPU
ConvNext-Base	ONNX	float	Snapdragon® X Elite	7.211 ms	176 - 176 MB	NPU
ConvNext-Base	ONNX	float	Snapdragon® 8 Gen 3 Mobile	5.32 ms	1 - 312 MB	NPU
ConvNext-Base	ONNX	float	Snapdragon® 8 Gen 1 Mobile	19.29 ms	1 - 301 MB	NPU
ConvNext-Base	ONNX	float	Qualcomm® Dragonwing™ IQ-8275	12.142 ms	0 - 4 MB	NPU
ConvNext-Base	ONNX	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	7.132 ms	0 - 227 MB	NPU
ConvNext-Base	ONNX	float	Qualcomm® QCS8450	19.29 ms	1 - 301 MB	NPU
ConvNext-Base	ONNX	float	Qualcomm® Dragonwing™ IQ-9075	10.939 ms	0 - 4 MB	NPU
ConvNext-Base	ONNX	float	Qualcomm® Dragonwing™ IQ-X7181	7.211 ms	176 - 176 MB	NPU
ConvNext-Base	ONNX	float	Qualcomm® Dragonwing™ Q-8750	4.158 ms	0 - 184 MB	NPU
ConvNext-Base	ONNX	float	Snapdragon® 8 Elite Mobile	4.158 ms	0 - 184 MB	NPU
ConvNext-Base	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	3.199 ms	1 - 188 MB	NPU
ConvNext-Base	ONNX	w8a16	Snapdragon® X2 Elite	2.383 ms	1 - 1 MB	NPU
ConvNext-Base	ONNX	w8a16	Snapdragon® X Elite	4.982 ms	91 - 91 MB	NPU
ConvNext-Base	ONNX	w8a16	Snapdragon® 8 Gen 3 Mobile	3.417 ms	0 - 262 MB	NPU
ConvNext-Base	ONNX	w8a16	Snapdragon® 8 Gen 1 Mobile	7.994 ms	0 - 262 MB	NPU
ConvNext-Base	ONNX	w8a16	Qualcomm® Dragonwing™ QCS6490	33.017 ms	0 - 3 MB	NPU
ConvNext-Base	ONNX	w8a16	Qualcomm® Dragonwing™ IQ-8275	4.886 ms	0 - 4 MB	NPU
ConvNext-Base	ONNX	w8a16	Qualcomm® Dragonwing™ QCS8550 (Proxy)	4.855 ms	0 - 103 MB	NPU
ConvNext-Base	ONNX	w8a16	Qualcomm® QCS8450	7.994 ms	0 - 262 MB	NPU
ConvNext-Base	ONNX	w8a16	Qualcomm® Dragonwing™ IQ-9075	4.884 ms	0 - 3 MB	NPU
ConvNext-Base	ONNX	w8a16	Qualcomm® Dragonwing™ IQ-X7181	4.982 ms	91 - 91 MB	NPU
ConvNext-Base	ONNX	w8a16	Qualcomm® Dragonwing™ Q-6690	57.965 ms	0 - 399 MB	NPU
ConvNext-Base	ONNX	w8a16	Qualcomm® Dragonwing™ Q-7790	6.91 ms	0 - 259 MB	NPU
ConvNext-Base	ONNX	w8a16	Qualcomm® Dragonwing™ Q-8750	2.78 ms	0 - 206 MB	NPU
ConvNext-Base	ONNX	w8a16	Snapdragon® 8 Elite Mobile	2.78 ms	0 - 206 MB	NPU
ConvNext-Base	ONNX	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	2.165 ms	0 - 226 MB	NPU
ConvNext-Base	ONNX	w8a16	Snapdragon® 7 Gen 4 Mobile	6.91 ms	0 - 259 MB	NPU
ConvNext-Base	QNN_DLC	float	Snapdragon® X2 Elite	4.401 ms	1 - 1 MB	NPU
ConvNext-Base	QNN_DLC	float	Snapdragon® X Elite	8.404 ms	1 - 1 MB	NPU
ConvNext-Base	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	5.851 ms	0 - 305 MB	NPU
ConvNext-Base	QNN_DLC	float	Snapdragon® 8 Gen 1 Mobile	20.34 ms	0 - 295 MB	NPU
ConvNext-Base	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-8275	12.71 ms	1 - 4 MB	NPU
ConvNext-Base	QNN_DLC	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	7.939 ms	1 - 362 MB	NPU
ConvNext-Base	QNN_DLC	float	Qualcomm® SA8775P	11.824 ms	1 - 181 MB	NPU
ConvNext-Base	QNN_DLC	float	Qualcomm® SA8650P	11.824 ms	1 - 181 MB	NPU
ConvNext-Base	QNN_DLC	float	Qualcomm® SA8255P	11.824 ms	1 - 181 MB	NPU
ConvNext-Base	QNN_DLC	float	Qualcomm® QCS8450	20.34 ms	0 - 295 MB	NPU
ConvNext-Base	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-9075	11.879 ms	1 - 3 MB	NPU
ConvNext-Base	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-X7181	8.404 ms	1 - 1 MB	NPU
ConvNext-Base	QNN_DLC	float	Qualcomm® Dragonwing™ Q-8750	4.526 ms	1 - 181 MB	NPU
ConvNext-Base	QNN_DLC	float	Qualcomm® SA7255P	41.896 ms	1 - 180 MB	NPU
ConvNext-Base	QNN_DLC	float	Qualcomm® SA8295P	19.676 ms	1 - 171 MB	NPU
ConvNext-Base	QNN_DLC	float	Snapdragon® 8 Elite Mobile	4.526 ms	1 - 181 MB	NPU
ConvNext-Base	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	3.554 ms	1 - 186 MB	NPU
ConvNext-Base	QNN_DLC	w8a16	Snapdragon® X2 Elite	3.114 ms	0 - 0 MB	NPU
ConvNext-Base	QNN_DLC	w8a16	Snapdragon® X Elite	6.259 ms	0 - 0 MB	NPU
ConvNext-Base	QNN_DLC	w8a16	Snapdragon® 8 Gen 3 Mobile	4.073 ms	0 - 249 MB	NPU
ConvNext-Base	QNN_DLC	w8a16	Snapdragon® 8 Gen 1 Mobile	9.115 ms	0 - 251 MB	NPU
ConvNext-Base	QNN_DLC	w8a16	Qualcomm® Dragonwing™ QCS6490	29.686 ms	0 - 2 MB	NPU
ConvNext-Base	QNN_DLC	w8a16	Qualcomm® Dragonwing™ IQ-8275	5.766 ms	0 - 3 MB	NPU
ConvNext-Base	QNN_DLC	w8a16	Qualcomm® Dragonwing™ QCS8550 (Proxy)	5.882 ms	0 - 2 MB	NPU
ConvNext-Base	QNN_DLC	w8a16	Qualcomm® SA8775P	6.189 ms	0 - 204 MB	NPU
ConvNext-Base	QNN_DLC	w8a16	Qualcomm® SA8650P	6.189 ms	0 - 204 MB	NPU
ConvNext-Base	QNN_DLC	w8a16	Qualcomm® SA8255P	6.189 ms	0 - 204 MB	NPU
ConvNext-Base	QNN_DLC	w8a16	Qualcomm® QCS8450	9.115 ms	0 - 251 MB	NPU
ConvNext-Base	QNN_DLC	w8a16	Qualcomm® Dragonwing™ IQ-9075	5.942 ms	0 - 2 MB	NPU
ConvNext-Base	QNN_DLC	w8a16	Qualcomm® Dragonwing™ IQ-X7181	6.259 ms	0 - 0 MB	NPU
ConvNext-Base	QNN_DLC	w8a16	Qualcomm® Dragonwing™ Q-6690	75.491 ms	0 - 399 MB	NPU
ConvNext-Base	QNN_DLC	w8a16	Qualcomm® Dragonwing™ Q-7790	7.732 ms	0 - 256 MB	NPU
ConvNext-Base	QNN_DLC	w8a16	Qualcomm® Dragonwing™ Q-8750	3.248 ms	0 - 194 MB	NPU
ConvNext-Base	QNN_DLC	w8a16	Qualcomm® SA7255P	14.621 ms	0 - 203 MB	NPU
ConvNext-Base	QNN_DLC	w8a16	Qualcomm® SA8295P	9.37 ms	0 - 204 MB	NPU
ConvNext-Base	QNN_DLC	w8a16	Snapdragon® 8 Elite Mobile	3.248 ms	0 - 194 MB	NPU
ConvNext-Base	QNN_DLC	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	2.563 ms	0 - 214 MB	NPU
ConvNext-Base	QNN_DLC	w8a16	Snapdragon® 7 Gen 4 Mobile	7.732 ms	0 - 256 MB	NPU
ConvNext-Base	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	5.452 ms	0 - 304 MB	NPU
ConvNext-Base	TFLITE	float	Snapdragon® 8 Gen 1 Mobile	19.594 ms	0 - 290 MB	NPU
ConvNext-Base	TFLITE	float	Qualcomm® Dragonwing™ IQ-8275	12.195 ms	0 - 178 MB	NPU
ConvNext-Base	TFLITE	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	7.256 ms	0 - 196 MB	NPU
ConvNext-Base	TFLITE	float	Qualcomm® SA8775P	11.089 ms	0 - 176 MB	NPU
ConvNext-Base	TFLITE	float	Qualcomm® SA8650P	11.089 ms	0 - 176 MB	NPU
ConvNext-Base	TFLITE	float	Qualcomm® SA8255P	11.089 ms	0 - 176 MB	NPU
ConvNext-Base	TFLITE	float	Qualcomm® QCS8450	19.594 ms	0 - 290 MB	NPU
ConvNext-Base	TFLITE	float	Qualcomm® Dragonwing™ IQ-9075	11.136 ms	0 - 177 MB	NPU
ConvNext-Base	TFLITE	float	Qualcomm® Dragonwing™ Q-8750	4.101 ms	0 - 179 MB	NPU
ConvNext-Base	TFLITE	float	Qualcomm® SA7255P	40.986 ms	0 - 175 MB	NPU
ConvNext-Base	TFLITE	float	Qualcomm® SA8295P	18.788 ms	0 - 160 MB	NPU
ConvNext-Base	TFLITE	float	Snapdragon® 8 Elite Mobile	4.101 ms	0 - 179 MB	NPU
ConvNext-Base	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	3.222 ms	0 - 182 MB	NPU

License

• The license for the original implementation of ConvNext-Base can be found here.

References

• A ConvNet for the 2020s
• Source Model Implementation

Community

• Join our AI Hub Slack community to collaborate, post questions and learn more about on-device AI.
• For questions or feedback please reach out to us.