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 1.27.1	Download
ONNX	w8a8	Universal	QAIRT 2.45, ONNX Runtime 1.27.1	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:

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

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
ResNet50	ONNX	float	Snapdragon® X2 Elite	0.98 ms	2 - 2 MB	NPU
ResNet50	ONNX	float	Snapdragon® X Elite	1.914 ms	49 - 49 MB	NPU
ResNet50	ONNX	float	Snapdragon® 8 Gen 3 Mobile	1.463 ms	0 - 81 MB	NPU
ResNet50	ONNX	float	Snapdragon® 8 Gen 1 Mobile	3.193 ms	0 - 67 MB	NPU
ResNet50	ONNX	float	Qualcomm® Dragonwing™ IQ-8275	3.333 ms	0 - 4 MB	NPU
ResNet50	ONNX	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	1.995 ms	0 - 57 MB	NPU
ResNet50	ONNX	float	Qualcomm® QCS8450	3.193 ms	0 - 67 MB	NPU
ResNet50	ONNX	float	Qualcomm® Dragonwing™ IQ-9075	3.015 ms	1 - 4 MB	NPU
ResNet50	ONNX	float	Qualcomm® Dragonwing™ IQ-X7181	1.914 ms	49 - 49 MB	NPU
ResNet50	ONNX	float	Qualcomm® Dragonwing™ Q-8750	1.215 ms	0 - 52 MB	NPU
ResNet50	ONNX	float	Snapdragon® 8 Elite Mobile	1.215 ms	0 - 52 MB	NPU
ResNet50	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	1.063 ms	0 - 54 MB	NPU
ResNet50	ONNX	w8a8	Snapdragon® X2 Elite	0.421 ms	1 - 1 MB	NPU
ResNet50	ONNX	w8a8	Snapdragon® X Elite	0.83 ms	25 - 25 MB	NPU
ResNet50	ONNX	w8a8	Snapdragon® 8 Gen 3 Mobile	0.685 ms	0 - 86 MB	NPU
ResNet50	ONNX	w8a8	Snapdragon® 8 Gen 1 Mobile	1.205 ms	0 - 85 MB	NPU
ResNet50	ONNX	w8a8	Qualcomm® Dragonwing™ QCS6490	3.502 ms	0 - 3 MB	NPU
ResNet50	ONNX	w8a8	Qualcomm® Dragonwing™ IQ-8275	0.996 ms	0 - 4 MB	NPU
ResNet50	ONNX	w8a8	Qualcomm® Dragonwing™ QCS8550 (Proxy)	0.907 ms	0 - 30 MB	NPU
ResNet50	ONNX	w8a8	Qualcomm® QCS8450	1.205 ms	0 - 85 MB	NPU
ResNet50	ONNX	w8a8	Qualcomm® Dragonwing™ IQ-9075	1.016 ms	0 - 3 MB	NPU
ResNet50	ONNX	w8a8	Qualcomm® Dragonwing™ IQ-X7181	0.83 ms	25 - 25 MB	NPU
ResNet50	ONNX	w8a8	Qualcomm® Dragonwing™ Q-6690	6.484 ms	0 - 171 MB	NPU
ResNet50	ONNX	w8a8	Qualcomm® Dragonwing™ Q-7790	1.202 ms	0 - 55 MB	NPU
ResNet50	ONNX	w8a8	Qualcomm® Dragonwing™ Q-8750	0.583 ms	0 - 48 MB	NPU
ResNet50	ONNX	w8a8	Snapdragon® 8 Elite Mobile	0.583 ms	0 - 48 MB	NPU
ResNet50	ONNX	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.559 ms	0 - 51 MB	NPU
ResNet50	ONNX	w8a8	Snapdragon® 7 Gen 4 Mobile	1.202 ms	0 - 55 MB	NPU
ResNet50	QNN_DLC	float	Snapdragon® X2 Elite	1.226 ms	1 - 1 MB	NPU
ResNet50	QNN_DLC	float	Snapdragon® X Elite	2.343 ms	1 - 1 MB	NPU
ResNet50	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	1.609 ms	0 - 77 MB	NPU
ResNet50	QNN_DLC	float	Snapdragon® 8 Gen 1 Mobile	3.739 ms	0 - 63 MB	NPU
ResNet50	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-8275	3.362 ms	1 - 4 MB	NPU
ResNet50	QNN_DLC	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	2.226 ms	0 - 11 MB	NPU
ResNet50	QNN_DLC	float	Qualcomm® SA8775P	3.396 ms	1 - 48 MB	NPU
ResNet50	QNN_DLC	float	Qualcomm® SA8650P	3.396 ms	1 - 48 MB	NPU
ResNet50	QNN_DLC	float	Qualcomm® SA8255P	3.396 ms	1 - 48 MB	NPU
ResNet50	QNN_DLC	float	Qualcomm® QCS8450	3.739 ms	0 - 63 MB	NPU
ResNet50	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-9075	3.187 ms	3 - 5 MB	NPU
ResNet50	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-X7181	2.343 ms	1 - 1 MB	NPU
ResNet50	QNN_DLC	float	Qualcomm® Dragonwing™ Q-8750	1.273 ms	0 - 48 MB	NPU
ResNet50	QNN_DLC	float	Qualcomm® SA7255P	10.724 ms	1 - 46 MB	NPU
ResNet50	QNN_DLC	float	Qualcomm® SA8295P	3.658 ms	1 - 29 MB	NPU
ResNet50	QNN_DLC	float	Snapdragon® 8 Elite Mobile	1.273 ms	0 - 48 MB	NPU
ResNet50	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	1.102 ms	1 - 49 MB	NPU
ResNet50	QNN_DLC	w8a8	Snapdragon® X2 Elite	0.453 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	0.682 ms	0 - 74 MB	NPU
ResNet50	QNN_DLC	w8a8	Snapdragon® 8 Gen 1 Mobile	1.195 ms	0 - 78 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® Dragonwing™ QCS6490	3.727 ms	0 - 2 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® Dragonwing™ IQ-8275	0.953 ms	0 - 3 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® Dragonwing™ QCS8550 (Proxy)	0.892 ms	0 - 2 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® SA8775P	1.091 ms	0 - 46 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® SA8650P	1.091 ms	0 - 46 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® SA8255P	1.091 ms	0 - 46 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® QCS8450	1.195 ms	0 - 78 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® Dragonwing™ IQ-9075	0.981 ms	2 - 4 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® Dragonwing™ IQ-X7181	0.933 ms	0 - 0 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® Dragonwing™ Q-6690	6.603 ms	0 - 163 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® Dragonwing™ Q-7790	1.221 ms	0 - 51 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® Dragonwing™ Q-8750	0.526 ms	0 - 42 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® SA7255P	2.029 ms	0 - 45 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® SA8295P	1.414 ms	0 - 42 MB	NPU
ResNet50	QNN_DLC	w8a8	Snapdragon® 8 Elite Mobile	0.526 ms	0 - 42 MB	NPU
ResNet50	QNN_DLC	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.486 ms	0 - 45 MB	NPU
ResNet50	QNN_DLC	w8a8	Snapdragon® 7 Gen 4 Mobile	1.221 ms	0 - 51 MB	NPU
ResNet50	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	1.584 ms	0 - 112 MB	NPU
ResNet50	TFLITE	float	Snapdragon® 8 Gen 1 Mobile	3.719 ms	0 - 97 MB	NPU
ResNet50	TFLITE	float	Qualcomm® Dragonwing™ IQ-8275	3.352 ms	0 - 53 MB	NPU
ResNet50	TFLITE	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	2.213 ms	0 - 2 MB	NPU
ResNet50	TFLITE	float	Qualcomm® SA8775P	3.356 ms	0 - 70 MB	NPU
ResNet50	TFLITE	float	Qualcomm® SA8650P	3.356 ms	0 - 70 MB	NPU
ResNet50	TFLITE	float	Qualcomm® SA8255P	3.356 ms	0 - 70 MB	NPU
ResNet50	TFLITE	float	Qualcomm® QCS8450	3.719 ms	0 - 97 MB	NPU
ResNet50	TFLITE	float	Qualcomm® Dragonwing™ IQ-9075	3.122 ms	0 - 52 MB	NPU
ResNet50	TFLITE	float	Qualcomm® Dragonwing™ Q-8750	1.251 ms	0 - 66 MB	NPU
ResNet50	TFLITE	float	Qualcomm® SA7255P	10.611 ms	0 - 69 MB	NPU
ResNet50	TFLITE	float	Qualcomm® SA8295P	3.637 ms	0 - 50 MB	NPU
ResNet50	TFLITE	float	Snapdragon® 8 Elite Mobile	1.251 ms	0 - 66 MB	NPU
ResNet50	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	1.092 ms	0 - 66 MB	NPU
ResNet50	TFLITE	w8a8	Snapdragon® 8 Gen 3 Mobile	0.585 ms	0 - 74 MB	NPU
ResNet50	TFLITE	w8a8	Snapdragon® 8 Gen 1 Mobile	1.053 ms	0 - 81 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® Dragonwing™ QCS6490	3.376 ms	0 - 27 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® Dragonwing™ IQ-8275	0.829 ms	0 - 27 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® Dragonwing™ QCS8550 (Proxy)	0.768 ms	0 - 11 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® SA8775P	0.966 ms	0 - 45 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® SA8650P	0.966 ms	0 - 45 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® SA8255P	0.966 ms	0 - 45 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® QCS8450	1.053 ms	0 - 81 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® Dragonwing™ IQ-9075	0.849 ms	0 - 27 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® Dragonwing™ Q-6690	6.116 ms	0 - 161 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® Dragonwing™ Q-7790	1.094 ms	0 - 49 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® Dragonwing™ Q-8750	0.48 ms	0 - 44 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® SA7255P	1.779 ms	0 - 42 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® SA8295P	1.285 ms	0 - 40 MB	NPU
ResNet50	TFLITE	w8a8	Snapdragon® 8 Elite Mobile	0.48 ms	0 - 44 MB	NPU
ResNet50	TFLITE	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.449 ms	0 - 43 MB	NPU
ResNet50	TFLITE	w8a8	Snapdragon® 7 Gen 4 Mobile	1.094 ms	0 - 49 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.