ResNet18: Optimized for Qualcomm Devices

ResNet18 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 ResNet18 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 ResNet18 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 ResNet18 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: 11.7M
• Model size (float): 44.6 MB
• Model size (w8a8): 11.3 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
ResNet18	ONNX	float	Snapdragon® X2 Elite	0.513 ms	212 - 212 MB	NPU
ResNet18	ONNX	float	Snapdragon® X Elite	1.022 ms	181 - 181 MB	NPU
ResNet18	ONNX	float	Snapdragon® 8 Gen 3 Mobile	0.779 ms	1 - 25 MB	NPU
ResNet18	ONNX	float	Snapdragon® 8 Gen 1 Mobile	1.839 ms	1 - 28 MB	NPU
ResNet18	ONNX	float	Qualcomm® QCS8550 (Proxy)	1.049 ms	1 - 4 MB	NPU
ResNet18	ONNX	float	Qualcomm® QCS8450	1.839 ms	1 - 28 MB	NPU
ResNet18	ONNX	float	Snapdragon® 8 Elite Mobile	0.616 ms	0 - 27 MB	NPU
ResNet18	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	0.552 ms	1 - 23 MB	NPU
ResNet18	ONNX	float	Qualcomm® QCS9075	1.803 ms	1 - 46 MB	NPU
ResNet18	ONNX	float	Qualcomm® QCS8750	0.616 ms	0 - 27 MB	NPU
ResNet18	ONNX	float	Qualcomm® QCS7181	1.022 ms	181 - 181 MB	NPU
ResNet18	ONNX	w8a8	Snapdragon® X2 Elite	0.265 ms	181 - 181 MB	NPU
ResNet18	ONNX	w8a8	Snapdragon® X Elite	0.494 ms	181 - 181 MB	NPU
ResNet18	ONNX	w8a8	Snapdragon® 8 Gen 3 Mobile	0.41 ms	0 - 38 MB	NPU
ResNet18	ONNX	w8a8	Snapdragon® 8 Gen 1 Mobile	0.748 ms	0 - 45 MB	NPU
ResNet18	ONNX	w8a8	Qualcomm® QCS6490	1.688 ms	0 - 45 MB	NPU
ResNet18	ONNX	w8a8	Qualcomm® QCS8550 (Proxy)	0.55 ms	0 - 79 MB	NPU
ResNet18	ONNX	w8a8	Qualcomm® QCS8450	0.748 ms	0 - 45 MB	NPU
ResNet18	ONNX	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.344 ms	0 - 25 MB	NPU
ResNet18	ONNX	w8a8	Snapdragon® 7 Gen 4 Mobile	0.65 ms	0 - 29 MB	NPU
ResNet18	ONNX	w8a8	Qualcomm® QCM6690	3.125 ms	0 - 31 MB	NPU
ResNet18	ONNX	w8a8	Qualcomm® QCS9075	0.667 ms	0 - 45 MB	NPU
ResNet18	ONNX	w8a8	Snapdragon® 8 Elite Mobile	0.374 ms	0 - 28 MB	NPU
ResNet18	ONNX	w8a8	Qualcomm® QCS7790	0.65 ms	0 - 29 MB	NPU
ResNet18	ONNX	w8a8	Qualcomm® QCS8750	0.374 ms	0 - 28 MB	NPU
ResNet18	ONNX	w8a8	Qualcomm® QCS7181	0.494 ms	181 - 181 MB	NPU
ResNet18	QNN_DLC	float	Snapdragon® X2 Elite	0.66 ms	1 - 1 MB	NPU
ResNet18	QNN_DLC	float	Snapdragon® X Elite	1.449 ms	1 - 1 MB	NPU
ResNet18	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	0.92 ms	0 - 35 MB	NPU
ResNet18	QNN_DLC	float	Snapdragon® 8 Gen 1 Mobile	2.423 ms	0 - 38 MB	NPU
ResNet18	QNN_DLC	float	Qualcomm® QCS8275	5.875 ms	1 - 24 MB	NPU
ResNet18	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	1.319 ms	1 - 2 MB	NPU
ResNet18	QNN_DLC	float	Qualcomm® QCS8450	2.423 ms	0 - 38 MB	NPU
ResNet18	QNN_DLC	float	Snapdragon® 8 Elite Mobile	0.706 ms	0 - 26 MB	NPU
ResNet18	QNN_DLC	float	Qualcomm® SA8295P	2.316 ms	0 - 20 MB	NPU
ResNet18	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	0.593 ms	1 - 29 MB	NPU
ResNet18	QNN_DLC	float	Qualcomm® SA7255P	5.875 ms	1 - 24 MB	NPU
ResNet18	QNN_DLC	float	Qualcomm® QCS9075	2.064 ms	1 - 3 MB	NPU
ResNet18	QNN_DLC	float	Qualcomm® QCS8750	0.706 ms	0 - 26 MB	NPU
ResNet18	QNN_DLC	float	Qualcomm® QCS7181	1.449 ms	1 - 1 MB	NPU
ResNet18	QNN_DLC	w8a8	Snapdragon® X2 Elite	0.302 ms	0 - 0 MB	NPU
ResNet18	QNN_DLC	w8a8	Snapdragon® X Elite	0.604 ms	0 - 0 MB	NPU
ResNet18	QNN_DLC	w8a8	Snapdragon® 8 Gen 3 Mobile	0.417 ms	0 - 48 MB	NPU
ResNet18	QNN_DLC	w8a8	Snapdragon® 8 Gen 1 Mobile	0.74 ms	0 - 53 MB	NPU
ResNet18	QNN_DLC	w8a8	Qualcomm® QCS6490	1.814 ms	0 - 2 MB	NPU
ResNet18	QNN_DLC	w8a8	Qualcomm® QCS8275	1.315 ms	0 - 25 MB	NPU
ResNet18	QNN_DLC	w8a8	Qualcomm® QCS8550 (Proxy)	0.543 ms	0 - 11 MB	NPU
ResNet18	QNN_DLC	w8a8	Qualcomm® QCS8450	0.74 ms	0 - 53 MB	NPU
ResNet18	QNN_DLC	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.266 ms	0 - 24 MB	NPU
ResNet18	QNN_DLC	w8a8	Snapdragon® 7 Gen 4 Mobile	0.678 ms	0 - 32 MB	NPU
ResNet18	QNN_DLC	w8a8	Qualcomm® QCM6690	3.238 ms	0 - 33 MB	NPU
ResNet18	QNN_DLC	w8a8	Qualcomm® QCS9075	0.618 ms	2 - 4 MB	NPU
ResNet18	QNN_DLC	w8a8	Qualcomm® SA7255P	1.315 ms	0 - 25 MB	NPU
ResNet18	QNN_DLC	w8a8	Snapdragon® 8 Elite Mobile	0.31 ms	0 - 27 MB	NPU
ResNet18	QNN_DLC	w8a8	Qualcomm® SA8295P	0.97 ms	0 - 22 MB	NPU
ResNet18	QNN_DLC	w8a8	Qualcomm® QCS7790	0.678 ms	0 - 32 MB	NPU
ResNet18	QNN_DLC	w8a8	Qualcomm® QCS8750	0.31 ms	0 - 27 MB	NPU
ResNet18	QNN_DLC	w8a8	Qualcomm® QCS7181	0.604 ms	0 - 0 MB	NPU
ResNet18	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	0.919 ms	0 - 55 MB	NPU
ResNet18	TFLITE	float	Snapdragon® 8 Gen 1 Mobile	2.456 ms	0 - 58 MB	NPU
ResNet18	TFLITE	float	Qualcomm® QCS8275	5.871 ms	0 - 26 MB	NPU
ResNet18	TFLITE	float	Qualcomm® QCS8550 (Proxy)	1.316 ms	0 - 2 MB	NPU
ResNet18	TFLITE	float	Qualcomm® SA8775P	10.831 ms	0 - 19 MB	GPU
ResNet18	TFLITE	float	Qualcomm® SA8650P	10.831 ms	0 - 19 MB	GPU
ResNet18	TFLITE	float	Qualcomm® SA8255P	10.831 ms	0 - 19 MB	GPU
ResNet18	TFLITE	float	Qualcomm® QCS8450	2.456 ms	0 - 58 MB	NPU
ResNet18	TFLITE	float	Snapdragon® 8 Elite Mobile	0.702 ms	0 - 26 MB	NPU
ResNet18	TFLITE	float	Qualcomm® SA8295P	2.294 ms	0 - 22 MB	NPU
ResNet18	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	0.587 ms	0 - 27 MB	NPU
ResNet18	TFLITE	float	Qualcomm® SA7255P	5.871 ms	0 - 26 MB	NPU
ResNet18	TFLITE	float	Qualcomm® QCS9075	2.037 ms	0 - 25 MB	NPU
ResNet18	TFLITE	float	Qualcomm® QCS8750	0.702 ms	0 - 26 MB	NPU
ResNet18	TFLITE	w8a8	Snapdragon® 8 Gen 3 Mobile	0.298 ms	0 - 53 MB	NPU
ResNet18	TFLITE	w8a8	Snapdragon® 8 Gen 1 Mobile	0.608 ms	0 - 48 MB	NPU
ResNet18	TFLITE	w8a8	Qualcomm® QCS6490	1.541 ms	0 - 13 MB	NPU
ResNet18	TFLITE	w8a8	Qualcomm® QCS8275	1.043 ms	0 - 23 MB	NPU
ResNet18	TFLITE	w8a8	Qualcomm® QCS8550 (Proxy)	0.389 ms	0 - 2 MB	NPU
ResNet18	TFLITE	w8a8	Qualcomm® SA8775P	11.13 ms	1 - 22 MB	GPU
ResNet18	TFLITE	w8a8	Qualcomm® SA8650P	11.13 ms	1 - 22 MB	GPU
ResNet18	TFLITE	w8a8	Qualcomm® SA8255P	11.13 ms	1 - 22 MB	GPU
ResNet18	TFLITE	w8a8	Qualcomm® QCS8450	0.608 ms	0 - 48 MB	NPU
ResNet18	TFLITE	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.229 ms	0 - 23 MB	NPU
ResNet18	TFLITE	w8a8	Snapdragon® 7 Gen 4 Mobile	0.549 ms	0 - 31 MB	NPU
ResNet18	TFLITE	w8a8	Qualcomm® QCM6690	2.792 ms	0 - 32 MB	NPU
ResNet18	TFLITE	w8a8	Qualcomm® QCS9075	0.476 ms	0 - 13 MB	NPU
ResNet18	TFLITE	w8a8	Qualcomm® SA7255P	1.043 ms	0 - 23 MB	NPU
ResNet18	TFLITE	w8a8	Snapdragon® 8 Elite Mobile	0.241 ms	0 - 27 MB	NPU
ResNet18	TFLITE	w8a8	Qualcomm® SA8295P	0.769 ms	0 - 21 MB	NPU
ResNet18	TFLITE	w8a8	Qualcomm® QCS7790	0.549 ms	0 - 31 MB	NPU
ResNet18	TFLITE	w8a8	Qualcomm® QCS8750	0.241 ms	0 - 27 MB	NPU

License

• The license for the original implementation of ResNet18 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.