ResNet101: Optimized for Qualcomm Devices

ResNet101 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 ResNet101 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 2.37, ONNX Runtime 1.23.0	Download
ONNX	w8a8	Universal	QAIRT 2.37, ONNX Runtime 1.23.0	Download
QNN_DLC	float	Universal	QAIRT 2.42	Download
QNN_DLC	w8a8	Universal	QAIRT 2.42	Download
TFLITE	float	Universal	QAIRT 2.42, TFLite 2.17.0	Download
TFLITE	w8a8	Universal	QAIRT 2.42, TFLite 2.17.0	Download

For more device-specific assets and performance metrics, visit ResNet101 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 ResNet101 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: 44.5M
• Model size (float): 170 MB
• Model size (w8a8): 43.9 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
ResNet101	ONNX	float	Snapdragon® X Elite	3.247 ms	86 - 86 MB	NPU
ResNet101	ONNX	float	Snapdragon® 8 Gen 3 Mobile	2.471 ms	0 - 167 MB	NPU
ResNet101	ONNX	float	Qualcomm® QCS8550 (Proxy)	3.378 ms	0 - 180 MB	NPU
ResNet101	ONNX	float	Qualcomm® QCS9075	5.374 ms	0 - 4 MB	NPU
ResNet101	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	2.029 ms	0 - 124 MB	NPU
ResNet101	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	1.767 ms	0 - 125 MB	NPU
ResNet101	ONNX	w8a8	Snapdragon® X Elite	1.283 ms	44 - 44 MB	NPU
ResNet101	ONNX	w8a8	Snapdragon® 8 Gen 3 Mobile	1.154 ms	0 - 208 MB	NPU
ResNet101	ONNX	w8a8	Qualcomm® QCS6490	56.63 ms	8 - 56 MB	CPU
ResNet101	ONNX	w8a8	Qualcomm® QCS8550 (Proxy)	1.499 ms	0 - 51 MB	NPU
ResNet101	ONNX	w8a8	Qualcomm® QCS9075	1.588 ms	0 - 3 MB	NPU
ResNet101	ONNX	w8a8	Qualcomm® QCM6690	41.586 ms	11 - 22 MB	CPU
ResNet101	ONNX	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	0.979 ms	0 - 145 MB	NPU
ResNet101	ONNX	w8a8	Snapdragon® 7 Gen 4 Mobile	38.871 ms	9 - 19 MB	CPU
ResNet101	ONNX	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.927 ms	0 - 146 MB	NPU
ResNet101	QNN_DLC	float	Snapdragon® X Elite	3.531 ms	1 - 1 MB	NPU
ResNet101	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	2.406 ms	0 - 125 MB	NPU
ResNet101	QNN_DLC	float	Qualcomm® QCS8275 (Proxy)	18.336 ms	1 - 66 MB	NPU
ResNet101	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	3.334 ms	1 - 3 MB	NPU
ResNet101	QNN_DLC	float	Qualcomm® SA8775P	5.468 ms	1 - 67 MB	NPU
ResNet101	QNN_DLC	float	Qualcomm® QCS9075	5.286 ms	1 - 3 MB	NPU
ResNet101	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	5.937 ms	0 - 91 MB	NPU
ResNet101	QNN_DLC	float	Qualcomm® SA7255P	18.336 ms	1 - 66 MB	NPU
ResNet101	QNN_DLC	float	Qualcomm® SA8295P	5.604 ms	1 - 41 MB	NPU
ResNet101	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	1.961 ms	1 - 69 MB	NPU
ResNet101	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	1.647 ms	1 - 70 MB	NPU
ResNet101	QNN_DLC	w8a8	Snapdragon® X Elite	1.271 ms	0 - 0 MB	NPU
ResNet101	QNN_DLC	w8a8	Snapdragon® 8 Gen 3 Mobile	0.986 ms	0 - 119 MB	NPU
ResNet101	QNN_DLC	w8a8	Qualcomm® QCS6490	4.554 ms	0 - 2 MB	NPU
ResNet101	QNN_DLC	w8a8	Qualcomm® QCS8275 (Proxy)	2.953 ms	0 - 70 MB	NPU
ResNet101	QNN_DLC	w8a8	Qualcomm® QCS8550 (Proxy)	1.268 ms	0 - 2 MB	NPU
ResNet101	QNN_DLC	w8a8	Qualcomm® SA8775P	5.196 ms	0 - 70 MB	NPU
ResNet101	QNN_DLC	w8a8	Qualcomm® QCS9075	1.31 ms	2 - 4 MB	NPU
ResNet101	QNN_DLC	w8a8	Qualcomm® QCM6690	11.7 ms	1 - 194 MB	NPU
ResNet101	QNN_DLC	w8a8	Qualcomm® QCS8450 (Proxy)	1.793 ms	0 - 121 MB	NPU
ResNet101	QNN_DLC	w8a8	Qualcomm® SA7255P	2.953 ms	0 - 70 MB	NPU
ResNet101	QNN_DLC	w8a8	Qualcomm® SA8295P	1.949 ms	0 - 68 MB	NPU
ResNet101	QNN_DLC	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	0.79 ms	0 - 70 MB	NPU
ResNet101	QNN_DLC	w8a8	Snapdragon® 7 Gen 4 Mobile	1.8 ms	0 - 80 MB	NPU
ResNet101	QNN_DLC	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.741 ms	0 - 72 MB	NPU
ResNet101	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	2.399 ms	0 - 192 MB	NPU
ResNet101	TFLITE	float	Qualcomm® QCS8275 (Proxy)	18.357 ms	0 - 128 MB	NPU
ResNet101	TFLITE	float	Qualcomm® QCS8550 (Proxy)	3.302 ms	0 - 3 MB	NPU
ResNet101	TFLITE	float	Qualcomm® SA8775P	5.434 ms	0 - 130 MB	NPU
ResNet101	TFLITE	float	Qualcomm® QCS9075	5.286 ms	0 - 88 MB	NPU
ResNet101	TFLITE	float	Qualcomm® QCS8450 (Proxy)	5.891 ms	0 - 158 MB	NPU
ResNet101	TFLITE	float	Qualcomm® SA7255P	18.357 ms	0 - 128 MB	NPU
ResNet101	TFLITE	float	Qualcomm® SA8295P	5.631 ms	0 - 97 MB	NPU
ResNet101	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	1.95 ms	0 - 129 MB	NPU
ResNet101	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	1.635 ms	0 - 132 MB	NPU
ResNet101	TFLITE	w8a8	Snapdragon® 8 Gen 3 Mobile	0.867 ms	0 - 134 MB	NPU
ResNet101	TFLITE	w8a8	Qualcomm® QCS6490	4.312 ms	0 - 45 MB	NPU
ResNet101	TFLITE	w8a8	Qualcomm® QCS8275 (Proxy)	2.731 ms	0 - 68 MB	NPU
ResNet101	TFLITE	w8a8	Qualcomm® QCS8550 (Proxy)	1.141 ms	0 - 1 MB	NPU
ResNet101	TFLITE	w8a8	Qualcomm® SA8775P	1.331 ms	0 - 70 MB	NPU
ResNet101	TFLITE	w8a8	Qualcomm® QCS9075	1.181 ms	0 - 45 MB	NPU
ResNet101	TFLITE	w8a8	Qualcomm® QCM6690	11.804 ms	0 - 195 MB	NPU
ResNet101	TFLITE	w8a8	Qualcomm® QCS8450 (Proxy)	1.661 ms	0 - 137 MB	NPU
ResNet101	TFLITE	w8a8	Qualcomm® SA7255P	2.731 ms	0 - 68 MB	NPU
ResNet101	TFLITE	w8a8	Qualcomm® SA8295P	1.771 ms	0 - 66 MB	NPU
ResNet101	TFLITE	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	0.742 ms	0 - 67 MB	NPU
ResNet101	TFLITE	w8a8	Snapdragon® 7 Gen 4 Mobile	1.666 ms	0 - 75 MB	NPU
ResNet101	TFLITE	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.702 ms	0 - 70 MB	NPU

License

• The license for the original implementation of ResNet101 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.