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 2.42, ONNX Runtime 1.24.3	Download
ONNX	w8a8	Universal	QAIRT 2.42, ONNX Runtime 1.24.3	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 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® 8 Elite Gen 5 Mobile	1.604 ms	1 - 77 MB	NPU
ResNet101	ONNX	float	Snapdragon® X2 Elite	1.627 ms	86 - 86 MB	NPU
ResNet101	ONNX	float	Snapdragon® X Elite	3.299 ms	85 - 85 MB	NPU
ResNet101	ONNX	float	Snapdragon® 8 Gen 3 Mobile	2.24 ms	0 - 132 MB	NPU
ResNet101	ONNX	float	Qualcomm® QCS8550 (Proxy)	3.079 ms	0 - 103 MB	NPU
ResNet101	ONNX	float	Qualcomm® QCS9075	5.143 ms	0 - 4 MB	NPU
ResNet101	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	1.872 ms	0 - 77 MB	NPU
ResNet101	ONNX	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.801 ms	0 - 76 MB	NPU
ResNet101	ONNX	w8a8	Snapdragon® X2 Elite	0.556 ms	43 - 43 MB	NPU
ResNet101	ONNX	w8a8	Snapdragon® X Elite	1.295 ms	43 - 43 MB	NPU
ResNet101	ONNX	w8a8	Snapdragon® 8 Gen 3 Mobile	0.977 ms	0 - 138 MB	NPU
ResNet101	ONNX	w8a8	Qualcomm® QCS6490	57.87 ms	10 - 57 MB	CPU
ResNet101	ONNX	w8a8	Qualcomm® QCS8550 (Proxy)	1.281 ms	0 - 54 MB	NPU
ResNet101	ONNX	w8a8	Qualcomm® QCS9075	1.336 ms	0 - 3 MB	NPU
ResNet101	ONNX	w8a8	Qualcomm® QCM6690	41.95 ms	1 - 12 MB	CPU
ResNet101	ONNX	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	0.84 ms	0 - 80 MB	NPU
ResNet101	ONNX	w8a8	Snapdragon® 7 Gen 4 Mobile	39.439 ms	13 - 24 MB	CPU
ResNet101	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	1.633 ms	0 - 67 MB	NPU
ResNet101	QNN_DLC	float	Snapdragon® X2 Elite	1.958 ms	1 - 1 MB	NPU
ResNet101	QNN_DLC	float	Snapdragon® X Elite	3.539 ms	1 - 1 MB	NPU
ResNet101	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	2.426 ms	0 - 121 MB	NPU
ResNet101	QNN_DLC	float	Qualcomm® QCS8275 (Proxy)	18.342 ms	1 - 65 MB	NPU
ResNet101	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	3.399 ms	1 - 239 MB	NPU
ResNet101	QNN_DLC	float	Qualcomm® SA8775P	5.454 ms	1 - 67 MB	NPU
ResNet101	QNN_DLC	float	Qualcomm® QCS9075	5.278 ms	3 - 5 MB	NPU
ResNet101	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	5.948 ms	0 - 90 MB	NPU
ResNet101	QNN_DLC	float	Qualcomm® SA7255P	18.342 ms	1 - 65 MB	NPU
ResNet101	QNN_DLC	float	Qualcomm® SA8295P	5.61 ms	1 - 39 MB	NPU
ResNet101	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	1.952 ms	0 - 65 MB	NPU
ResNet101	QNN_DLC	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.729 ms	0 - 76 MB	NPU
ResNet101	QNN_DLC	w8a8	Snapdragon® X2 Elite	0.647 ms	0 - 0 MB	NPU
ResNet101	QNN_DLC	w8a8	Snapdragon® X Elite	1.278 ms	0 - 0 MB	NPU
ResNet101	QNN_DLC	w8a8	Snapdragon® 8 Gen 3 Mobile	0.974 ms	0 - 122 MB	NPU
ResNet101	QNN_DLC	w8a8	Qualcomm® QCS6490	4.471 ms	0 - 2 MB	NPU
ResNet101	QNN_DLC	w8a8	Qualcomm® QCS8275 (Proxy)	3.02 ms	0 - 72 MB	NPU
ResNet101	QNN_DLC	w8a8	Qualcomm® QCS8550 (Proxy)	1.286 ms	0 - 2 MB	NPU
ResNet101	QNN_DLC	w8a8	Qualcomm® SA8775P	1.525 ms	0 - 75 MB	NPU
ResNet101	QNN_DLC	w8a8	Qualcomm® QCS9075	1.322 ms	0 - 2 MB	NPU
ResNet101	QNN_DLC	w8a8	Qualcomm® QCM6690	11.71 ms	0 - 195 MB	NPU
ResNet101	QNN_DLC	w8a8	Qualcomm® QCS8450 (Proxy)	1.781 ms	0 - 124 MB	NPU
ResNet101	QNN_DLC	w8a8	Qualcomm® SA7255P	3.02 ms	0 - 72 MB	NPU
ResNet101	QNN_DLC	w8a8	Qualcomm® SA8295P	1.945 ms	0 - 71 MB	NPU
ResNet101	QNN_DLC	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	0.81 ms	0 - 74 MB	NPU
ResNet101	QNN_DLC	w8a8	Snapdragon® 7 Gen 4 Mobile	1.795 ms	0 - 83 MB	NPU
ResNet101	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	1.62 ms	0 - 112 MB	NPU
ResNet101	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	2.389 ms	0 - 178 MB	NPU
ResNet101	TFLITE	float	Qualcomm® QCS8275 (Proxy)	18.248 ms	0 - 108 MB	NPU
ResNet101	TFLITE	float	Qualcomm® QCS8550 (Proxy)	3.305 ms	0 - 3 MB	NPU
ResNet101	TFLITE	float	Qualcomm® SA8775P	5.428 ms	0 - 112 MB	NPU
ResNet101	TFLITE	float	Qualcomm® QCS9075	5.278 ms	0 - 88 MB	NPU
ResNet101	TFLITE	float	Qualcomm® QCS8450 (Proxy)	5.926 ms	0 - 145 MB	NPU
ResNet101	TFLITE	float	Qualcomm® SA7255P	18.248 ms	0 - 108 MB	NPU
ResNet101	TFLITE	float	Qualcomm® SA8295P	5.57 ms	0 - 82 MB	NPU
ResNet101	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	1.926 ms	0 - 112 MB	NPU
ResNet101	TFLITE	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.701 ms	0 - 74 MB	NPU
ResNet101	TFLITE	w8a8	Snapdragon® 8 Gen 3 Mobile	0.865 ms	0 - 134 MB	NPU
ResNet101	TFLITE	w8a8	Qualcomm® QCS6490	4.316 ms	0 - 45 MB	NPU
ResNet101	TFLITE	w8a8	Qualcomm® QCS8275 (Proxy)	2.746 ms	0 - 71 MB	NPU
ResNet101	TFLITE	w8a8	Qualcomm® QCS8550 (Proxy)	1.14 ms	0 - 2 MB	NPU
ResNet101	TFLITE	w8a8	Qualcomm® SA8775P	1.41 ms	0 - 74 MB	NPU
ResNet101	TFLITE	w8a8	Qualcomm® QCS9075	1.17 ms	0 - 45 MB	NPU
ResNet101	TFLITE	w8a8	Qualcomm® QCM6690	11.227 ms	0 - 191 MB	NPU
ResNet101	TFLITE	w8a8	Qualcomm® QCS8450 (Proxy)	1.64 ms	0 - 134 MB	NPU
ResNet101	TFLITE	w8a8	Qualcomm® SA7255P	2.746 ms	0 - 71 MB	NPU
ResNet101	TFLITE	w8a8	Qualcomm® SA8295P	1.801 ms	0 - 68 MB	NPU
ResNet101	TFLITE	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	0.742 ms	0 - 70 MB	NPU
ResNet101	TFLITE	w8a8	Snapdragon® 7 Gen 4 Mobile	1.668 ms	0 - 78 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.