ResNet50: Optimized for Qualcomm Devices

ResNet50 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 ResNet50 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 1.25.0	Download
ONNX	w8a8	Universal	QAIRT 2.45, ONNX Runtime 1.25.0	Download
QNN_DLC	float	Universal	QAIRT 2.45	Download
QNN_DLC	w8a8	Universal	QAIRT 2.45	Download
TFLITE	float	Universal	QAIRT 2.45	Download
TFLITE	w8a8	Universal	QAIRT 2.45	Download

For more device-specific assets and performance metrics, visit ResNet50 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 ResNet50 on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.image_classification

Model Stats:

• Model checkpoint: Imagenet
• Input resolution: 224x224
• Number of parameters: 25.5M
• Model size (float): 97.4 MB
• Model size (w8a8): 25.1 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
ResNet50	ONNX	float	Snapdragon® X2 Elite	0.982 ms	180 - 180 MB	NPU
ResNet50	ONNX	float	Snapdragon® X Elite	1.914 ms	181 - 181 MB	NPU
ResNet50	ONNX	float	Snapdragon® 8 Gen 3 Mobile	1.471 ms	1 - 71 MB	NPU
ResNet50	ONNX	float	Snapdragon® 8 Gen 1 Mobile	3.18 ms	1 - 54 MB	NPU
ResNet50	ONNX	float	Qualcomm® QCS8550 (Proxy)	1.979 ms	0 - 60 MB	NPU
ResNet50	ONNX	float	Qualcomm® QCS8450	3.18 ms	1 - 54 MB	NPU
ResNet50	ONNX	float	Snapdragon® 8 Elite Mobile	1.218 ms	0 - 54 MB	NPU
ResNet50	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	1.057 ms	1 - 49 MB	NPU
ResNet50	ONNX	float	Qualcomm® QCS9075	3.168 ms	1 - 46 MB	NPU
ResNet50	ONNX	float	Qualcomm® QCS8750	1.218 ms	0 - 54 MB	NPU
ResNet50	ONNX	float	Qualcomm® QCS7181	1.914 ms	181 - 181 MB	NPU
ResNet50	ONNX	w8a8	Snapdragon® X2 Elite	0.415 ms	181 - 181 MB	NPU
ResNet50	ONNX	w8a8	Snapdragon® X Elite	0.831 ms	181 - 181 MB	NPU
ResNet50	ONNX	w8a8	Snapdragon® 8 Gen 3 Mobile	0.687 ms	0 - 75 MB	NPU
ResNet50	ONNX	w8a8	Snapdragon® 8 Gen 1 Mobile	1.216 ms	0 - 75 MB	NPU
ResNet50	ONNX	w8a8	Qualcomm® QCS6490	2.884 ms	0 - 45 MB	NPU
ResNet50	ONNX	w8a8	Qualcomm® QCS8550 (Proxy)	0.91 ms	0 - 35 MB	NPU
ResNet50	ONNX	w8a8	Qualcomm® QCS8450	1.216 ms	0 - 75 MB	NPU
ResNet50	ONNX	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.558 ms	0 - 46 MB	NPU
ResNet50	ONNX	w8a8	Snapdragon® 7 Gen 4 Mobile	1.202 ms	0 - 51 MB	NPU
ResNet50	ONNX	w8a8	Qualcomm® QCM6690	6.481 ms	0 - 169 MB	NPU
ResNet50	ONNX	w8a8	Qualcomm® QCS9075	0.991 ms	0 - 45 MB	NPU
ResNet50	ONNX	w8a8	Snapdragon® 8 Elite Mobile	0.587 ms	0 - 50 MB	NPU
ResNet50	ONNX	w8a8	Qualcomm® QCS7790	1.202 ms	0 - 51 MB	NPU
ResNet50	ONNX	w8a8	Qualcomm® QCS8750	0.587 ms	0 - 50 MB	NPU
ResNet50	ONNX	w8a8	Qualcomm® QCS7181	0.831 ms	181 - 181 MB	NPU
ResNet50	QNN_DLC	float	Snapdragon® X2 Elite	1.219 ms	1 - 1 MB	NPU
ResNet50	QNN_DLC	float	Snapdragon® X Elite	2.327 ms	1 - 1 MB	NPU
ResNet50	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	1.621 ms	0 - 76 MB	NPU
ResNet50	QNN_DLC	float	Snapdragon® 8 Gen 1 Mobile	3.727 ms	0 - 61 MB	NPU
ResNet50	QNN_DLC	float	Qualcomm® QCS8275	10.725 ms	1 - 46 MB	NPU
ResNet50	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	2.242 ms	1 - 136 MB	NPU
ResNet50	QNN_DLC	float	Qualcomm® QCS8450	3.727 ms	0 - 61 MB	NPU
ResNet50	QNN_DLC	float	Snapdragon® 8 Elite Mobile	1.262 ms	1 - 51 MB	NPU
ResNet50	QNN_DLC	float	Qualcomm® SA8295P	3.668 ms	0 - 28 MB	NPU
ResNet50	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	1.09 ms	1 - 48 MB	NPU
ResNet50	QNN_DLC	float	Qualcomm® SA7255P	10.725 ms	1 - 46 MB	NPU
ResNet50	QNN_DLC	float	Qualcomm® QCS9075	3.298 ms	3 - 5 MB	NPU
ResNet50	QNN_DLC	float	Qualcomm® QCS8750	1.262 ms	1 - 51 MB	NPU
ResNet50	QNN_DLC	float	Qualcomm® QCS7181	2.327 ms	1 - 1 MB	NPU
ResNet50	QNN_DLC	w8a8	Snapdragon® X2 Elite	0.468 ms	0 - 0 MB	NPU
ResNet50	QNN_DLC	w8a8	Snapdragon® X Elite	0.933 ms	0 - 0 MB	NPU
ResNet50	QNN_DLC	w8a8	Snapdragon® 8 Gen 3 Mobile	0.683 ms	0 - 77 MB	NPU
ResNet50	QNN_DLC	w8a8	Snapdragon® 8 Gen 1 Mobile	1.193 ms	0 - 78 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® QCS6490	2.986 ms	0 - 2 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® QCS8275	2.015 ms	0 - 45 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® QCS8550 (Proxy)	0.906 ms	0 - 6 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® QCS8450	1.193 ms	0 - 78 MB	NPU
ResNet50	QNN_DLC	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.488 ms	0 - 44 MB	NPU
ResNet50	QNN_DLC	w8a8	Snapdragon® 7 Gen 4 Mobile	1.223 ms	0 - 51 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® QCM6690	6.616 ms	0 - 164 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® QCS9075	0.968 ms	0 - 2 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® SA7255P	2.015 ms	0 - 45 MB	NPU
ResNet50	QNN_DLC	w8a8	Snapdragon® 8 Elite Mobile	0.525 ms	0 - 42 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® SA8295P	1.442 ms	0 - 42 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® QCS7790	1.223 ms	0 - 51 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® QCS8750	0.525 ms	0 - 42 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® QCS7181	0.933 ms	0 - 0 MB	NPU
ResNet50	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	1.624 ms	0 - 115 MB	NPU
ResNet50	TFLITE	float	Snapdragon® 8 Gen 1 Mobile	3.745 ms	0 - 95 MB	NPU
ResNet50	TFLITE	float	Qualcomm® QCS8275	10.663 ms	0 - 69 MB	NPU
ResNet50	TFLITE	float	Qualcomm® QCS8550 (Proxy)	2.206 ms	0 - 4 MB	NPU
ResNet50	TFLITE	float	Qualcomm® SA8775P	20.779 ms	0 - 21 MB	GPU
ResNet50	TFLITE	float	Qualcomm® SA8650P	20.779 ms	0 - 21 MB	GPU
ResNet50	TFLITE	float	Qualcomm® SA8255P	20.779 ms	0 - 21 MB	GPU
ResNet50	TFLITE	float	Qualcomm® QCS8450	3.745 ms	0 - 95 MB	NPU
ResNet50	TFLITE	float	Snapdragon® 8 Elite Mobile	1.249 ms	0 - 67 MB	NPU
ResNet50	TFLITE	float	Qualcomm® SA8295P	3.638 ms	0 - 50 MB	NPU
ResNet50	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	1.092 ms	0 - 66 MB	NPU
ResNet50	TFLITE	float	Qualcomm® SA7255P	10.663 ms	0 - 69 MB	NPU
ResNet50	TFLITE	float	Qualcomm® QCS9075	3.338 ms	0 - 52 MB	NPU
ResNet50	TFLITE	float	Qualcomm® QCS8750	1.249 ms	0 - 67 MB	NPU
ResNet50	TFLITE	w8a8	Snapdragon® 8 Gen 3 Mobile	0.576 ms	0 - 74 MB	NPU
ResNet50	TFLITE	w8a8	Snapdragon® 8 Gen 1 Mobile	1.05 ms	0 - 80 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® QCS6490	2.629 ms	0 - 27 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® QCS8275	1.775 ms	0 - 42 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® QCS8550 (Proxy)	0.767 ms	0 - 3 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® SA8775P	21.203 ms	1 - 25 MB	GPU
ResNet50	TFLITE	w8a8	Qualcomm® SA8650P	21.203 ms	1 - 25 MB	GPU
ResNet50	TFLITE	w8a8	Qualcomm® SA8255P	21.203 ms	1 - 25 MB	GPU
ResNet50	TFLITE	w8a8	Qualcomm® QCS8450	1.05 ms	0 - 80 MB	NPU
ResNet50	TFLITE	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.448 ms	0 - 42 MB	NPU
ResNet50	TFLITE	w8a8	Snapdragon® 7 Gen 4 Mobile	1.095 ms	0 - 49 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® QCM6690	6.164 ms	0 - 161 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® QCS9075	0.821 ms	0 - 27 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® SA7255P	1.775 ms	0 - 42 MB	NPU
ResNet50	TFLITE	w8a8	Snapdragon® 8 Elite Mobile	0.495 ms	0 - 44 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® SA8295P	1.295 ms	0 - 39 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® QCS7790	1.095 ms	0 - 49 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® QCS8750	0.495 ms	0 - 44 MB	NPU

License

• The license for the original implementation of ResNet50 can be found here.

References

• Deep Residual Learning for Image Recognition
• 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.