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 2.42, ONNX Runtime 1.24.1	Download
ONNX	w8a8	Universal	QAIRT 2.42, ONNX Runtime 1.24.1	Download
QNN_DLC	float	Universal	QAIRT 2.43	Download
QNN_DLC	w8a8	Universal	QAIRT 2.43	Download
TFLITE	float	Universal	QAIRT 2.43, TFLite 2.17.0	Download
TFLITE	w8a8	Universal	QAIRT 2.43, TFLite 2.17.0	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	0.998 ms	49 - 49 MB	NPU
ResNet50	ONNX	float	Snapdragon® X Elite	2.088 ms	49 - 49 MB	NPU
ResNet50	ONNX	float	Snapdragon® 8 Gen 3 Mobile	1.477 ms	0 - 86 MB	NPU
ResNet50	ONNX	float	Qualcomm® QCS8550 (Proxy)	1.991 ms	0 - 4 MB	NPU
ResNet50	ONNX	float	Qualcomm® QCS9075	3.172 ms	1 - 4 MB	NPU
ResNet50	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	1.246 ms	0 - 44 MB	NPU
ResNet50	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	1.059 ms	1 - 75 MB	NPU
ResNet50	ONNX	w8a8	Snapdragon® X2 Elite	0.417 ms	25 - 25 MB	NPU
ResNet50	ONNX	w8a8	Snapdragon® X Elite	0.97 ms	25 - 25 MB	NPU
ResNet50	ONNX	w8a8	Snapdragon® 8 Gen 3 Mobile	0.676 ms	0 - 79 MB	NPU
ResNet50	ONNX	w8a8	Qualcomm® QCS6490	31.854 ms	9 - 28 MB	CPU
ResNet50	ONNX	w8a8	Qualcomm® QCS8550 (Proxy)	0.897 ms	0 - 31 MB	NPU
ResNet50	ONNX	w8a8	Qualcomm® QCS9075	0.968 ms	0 - 3 MB	NPU
ResNet50	ONNX	w8a8	Qualcomm® QCM6690	23.355 ms	7 - 15 MB	CPU
ResNet50	ONNX	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	0.578 ms	0 - 42 MB	NPU
ResNet50	ONNX	w8a8	Snapdragon® 7 Gen 4 Mobile	17.932 ms	6 - 15 MB	CPU
ResNet50	ONNX	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.554 ms	0 - 44 MB	NPU
ResNet50	QNN_DLC	float	Snapdragon® X2 Elite	1.254 ms	1 - 1 MB	NPU
ResNet50	QNN_DLC	float	Snapdragon® X Elite	2.317 ms	1 - 1 MB	NPU
ResNet50	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	1.607 ms	0 - 77 MB	NPU
ResNet50	QNN_DLC	float	Qualcomm® QCS8275 (Proxy)	10.626 ms	1 - 46 MB	NPU
ResNet50	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	2.192 ms	1 - 2 MB	NPU
ResNet50	QNN_DLC	float	Qualcomm® SA8775P	3.361 ms	1 - 47 MB	NPU
ResNet50	QNN_DLC	float	Qualcomm® QCS9075	3.35 ms	3 - 5 MB	NPU
ResNet50	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	3.761 ms	0 - 62 MB	NPU
ResNet50	QNN_DLC	float	Qualcomm® SA7255P	10.626 ms	1 - 46 MB	NPU
ResNet50	QNN_DLC	float	Qualcomm® SA8295P	3.668 ms	1 - 30 MB	NPU
ResNet50	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	1.262 ms	1 - 48 MB	NPU
ResNet50	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	1.084 ms	1 - 51 MB	NPU
ResNet50	QNN_DLC	w8a8	Snapdragon® X2 Elite	0.467 ms	0 - 0 MB	NPU
ResNet50	QNN_DLC	w8a8	Snapdragon® X Elite	0.962 ms	0 - 0 MB	NPU
ResNet50	QNN_DLC	w8a8	Snapdragon® 8 Gen 3 Mobile	0.683 ms	0 - 70 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® QCS6490	3.008 ms	0 - 2 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® QCS8275 (Proxy)	1.976 ms	0 - 40 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® QCS8550 (Proxy)	0.895 ms	0 - 2 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® SA8775P	1.096 ms	0 - 44 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® QCS9075	0.972 ms	0 - 2 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® QCM6690	6.477 ms	0 - 161 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® QCS8450 (Proxy)	1.194 ms	0 - 72 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® SA7255P	1.976 ms	0 - 40 MB	NPU
ResNet50	QNN_DLC	w8a8	Qualcomm® SA8295P	1.436 ms	0 - 38 MB	NPU
ResNet50	QNN_DLC	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	0.531 ms	0 - 39 MB	NPU
ResNet50	QNN_DLC	w8a8	Snapdragon® 7 Gen 4 Mobile	1.238 ms	0 - 48 MB	NPU
ResNet50	QNN_DLC	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.481 ms	0 - 42 MB	NPU
ResNet50	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	1.584 ms	0 - 120 MB	NPU
ResNet50	TFLITE	float	Qualcomm® QCS8275 (Proxy)	10.491 ms	0 - 80 MB	NPU
ResNet50	TFLITE	float	Qualcomm® QCS8550 (Proxy)	2.2 ms	0 - 2 MB	NPU
ResNet50	TFLITE	float	Qualcomm® SA8775P	3.39 ms	0 - 81 MB	NPU
ResNet50	TFLITE	float	Qualcomm® QCS9075	3.35 ms	0 - 52 MB	NPU
ResNet50	TFLITE	float	Qualcomm® QCS8450 (Proxy)	3.723 ms	0 - 105 MB	NPU
ResNet50	TFLITE	float	Qualcomm® SA7255P	10.491 ms	0 - 80 MB	NPU
ResNet50	TFLITE	float	Qualcomm® SA8295P	3.614 ms	0 - 61 MB	NPU
ResNet50	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	1.26 ms	0 - 85 MB	NPU
ResNet50	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	1.084 ms	0 - 83 MB	NPU
ResNet50	TFLITE	w8a8	Snapdragon® 8 Gen 3 Mobile	0.586 ms	0 - 78 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® QCS6490	2.581 ms	0 - 27 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® QCS8275 (Proxy)	1.746 ms	0 - 40 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® QCS8550 (Proxy)	0.759 ms	0 - 145 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® SA8775P	0.977 ms	0 - 43 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® QCS9075	0.826 ms	0 - 27 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® QCM6690	6.208 ms	0 - 159 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® QCS8450 (Proxy)	1.052 ms	0 - 73 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® SA7255P	1.746 ms	0 - 40 MB	NPU
ResNet50	TFLITE	w8a8	Qualcomm® SA8295P	1.28 ms	0 - 37 MB	NPU
ResNet50	TFLITE	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	0.481 ms	0 - 38 MB	NPU
ResNet50	TFLITE	w8a8	Snapdragon® 7 Gen 4 Mobile	1.086 ms	0 - 46 MB	NPU
ResNet50	TFLITE	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.446 ms	0 - 42 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.