ResNeXt50: Optimized for Qualcomm Devices

ResNeXt50 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 ResNeXt50 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 ResNeXt50 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 ResNeXt50 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: 25.0M
• Model size (float): 95.4 MB
• Model size (w8a8): 24.8 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
ResNeXt50	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	1.136 ms	1 - 85 MB	NPU
ResNeXt50	ONNX	float	Snapdragon® X2 Elite	1.102 ms	50 - 50 MB	NPU
ResNeXt50	ONNX	float	Snapdragon® X Elite	2.416 ms	50 - 50 MB	NPU
ResNeXt50	ONNX	float	Snapdragon® 8 Gen 3 Mobile	1.688 ms	0 - 151 MB	NPU
ResNeXt50	ONNX	float	Qualcomm® QCS8550 (Proxy)	2.264 ms	0 - 20 MB	NPU
ResNeXt50	ONNX	float	Qualcomm® QCS9075	3.477 ms	0 - 4 MB	NPU
ResNeXt50	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	1.382 ms	0 - 84 MB	NPU
ResNeXt50	ONNX	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.599 ms	0 - 78 MB	NPU
ResNeXt50	ONNX	w8a8	Snapdragon® X2 Elite	0.51 ms	25 - 25 MB	NPU
ResNeXt50	ONNX	w8a8	Snapdragon® X Elite	1.262 ms	25 - 25 MB	NPU
ResNeXt50	ONNX	w8a8	Snapdragon® 8 Gen 3 Mobile	0.832 ms	0 - 106 MB	NPU
ResNeXt50	ONNX	w8a8	Qualcomm® QCS6490	46.224 ms	6 - 25 MB	CPU
ResNeXt50	ONNX	w8a8	Qualcomm® QCS8550 (Proxy)	1.099 ms	0 - 31 MB	NPU
ResNeXt50	ONNX	w8a8	Qualcomm® QCS9075	1.228 ms	0 - 3 MB	NPU
ResNeXt50	ONNX	w8a8	Qualcomm® QCM6690	26.204 ms	4 - 12 MB	CPU
ResNeXt50	ONNX	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	0.702 ms	0 - 78 MB	NPU
ResNeXt50	ONNX	w8a8	Snapdragon® 7 Gen 4 Mobile	20.763 ms	4 - 13 MB	CPU
ResNeXt50	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	1.185 ms	1 - 77 MB	NPU
ResNeXt50	QNN_DLC	float	Snapdragon® X2 Elite	1.449 ms	1 - 1 MB	NPU
ResNeXt50	QNN_DLC	float	Snapdragon® X Elite	2.674 ms	1 - 1 MB	NPU
ResNeXt50	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	1.792 ms	0 - 142 MB	NPU
ResNeXt50	QNN_DLC	float	Qualcomm® QCS8275 (Proxy)	11.868 ms	1 - 76 MB	NPU
ResNeXt50	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	2.528 ms	1 - 2 MB	NPU
ResNeXt50	QNN_DLC	float	Qualcomm® SA8775P	3.812 ms	1 - 76 MB	NPU
ResNeXt50	QNN_DLC	float	Qualcomm® QCS9075	3.658 ms	1 - 3 MB	NPU
ResNeXt50	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	4.055 ms	0 - 115 MB	NPU
ResNeXt50	QNN_DLC	float	Qualcomm® SA7255P	11.868 ms	1 - 76 MB	NPU
ResNeXt50	QNN_DLC	float	Qualcomm® SA8295P	4.097 ms	0 - 53 MB	NPU
ResNeXt50	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	1.457 ms	0 - 77 MB	NPU
ResNeXt50	QNN_DLC	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.541 ms	0 - 72 MB	NPU
ResNeXt50	QNN_DLC	w8a8	Snapdragon® X2 Elite	0.675 ms	0 - 0 MB	NPU
ResNeXt50	QNN_DLC	w8a8	Snapdragon® X Elite	1.225 ms	0 - 0 MB	NPU
ResNeXt50	QNN_DLC	w8a8	Snapdragon® 8 Gen 3 Mobile	0.805 ms	0 - 96 MB	NPU
ResNeXt50	QNN_DLC	w8a8	Qualcomm® QCS6490	3.115 ms	0 - 2 MB	NPU
ResNeXt50	QNN_DLC	w8a8	Qualcomm® QCS8275 (Proxy)	2.476 ms	0 - 69 MB	NPU
ResNeXt50	QNN_DLC	w8a8	Qualcomm® QCS8550 (Proxy)	1.102 ms	0 - 2 MB	NPU
ResNeXt50	QNN_DLC	w8a8	Qualcomm® SA8775P	1.485 ms	0 - 70 MB	NPU
ResNeXt50	QNN_DLC	w8a8	Qualcomm® QCS9075	1.234 ms	2 - 4 MB	NPU
ResNeXt50	QNN_DLC	w8a8	Qualcomm® QCM6690	7.615 ms	0 - 192 MB	NPU
ResNeXt50	QNN_DLC	w8a8	Qualcomm® QCS8450 (Proxy)	1.447 ms	0 - 98 MB	NPU
ResNeXt50	QNN_DLC	w8a8	Qualcomm® SA7255P	2.476 ms	0 - 69 MB	NPU
ResNeXt50	QNN_DLC	w8a8	Qualcomm® SA8295P	1.773 ms	0 - 66 MB	NPU
ResNeXt50	QNN_DLC	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	0.624 ms	0 - 71 MB	NPU
ResNeXt50	QNN_DLC	w8a8	Snapdragon® 7 Gen 4 Mobile	1.396 ms	0 - 74 MB	NPU
ResNeXt50	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	1.198 ms	0 - 126 MB	NPU
ResNeXt50	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	1.768 ms	0 - 191 MB	NPU
ResNeXt50	TFLITE	float	Qualcomm® QCS8275 (Proxy)	11.945 ms	0 - 123 MB	NPU
ResNeXt50	TFLITE	float	Qualcomm® QCS8550 (Proxy)	2.467 ms	0 - 2 MB	NPU
ResNeXt50	TFLITE	float	Qualcomm® SA8775P	3.858 ms	0 - 124 MB	NPU
ResNeXt50	TFLITE	float	Qualcomm® QCS9075	3.71 ms	0 - 52 MB	NPU
ResNeXt50	TFLITE	float	Qualcomm® QCS8450 (Proxy)	3.98 ms	0 - 162 MB	NPU
ResNeXt50	TFLITE	float	Qualcomm® SA7255P	11.945 ms	0 - 123 MB	NPU
ResNeXt50	TFLITE	float	Qualcomm® SA8295P	4.17 ms	0 - 105 MB	NPU
ResNeXt50	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	1.422 ms	0 - 123 MB	NPU
ResNeXt50	TFLITE	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.483 ms	0 - 69 MB	NPU
ResNeXt50	TFLITE	w8a8	Snapdragon® 8 Gen 3 Mobile	0.663 ms	0 - 95 MB	NPU
ResNeXt50	TFLITE	w8a8	Qualcomm® QCS6490	2.815 ms	0 - 27 MB	NPU
ResNeXt50	TFLITE	w8a8	Qualcomm® QCS8275 (Proxy)	2.223 ms	0 - 67 MB	NPU
ResNeXt50	TFLITE	w8a8	Qualcomm® QCS8550 (Proxy)	0.914 ms	0 - 2 MB	NPU
ResNeXt50	TFLITE	w8a8	Qualcomm® SA8775P	1.295 ms	0 - 68 MB	NPU
ResNeXt50	TFLITE	w8a8	Qualcomm® QCS9075	1.01 ms	0 - 27 MB	NPU
ResNeXt50	TFLITE	w8a8	Qualcomm® QCM6690	7.327 ms	0 - 191 MB	NPU
ResNeXt50	TFLITE	w8a8	Qualcomm® QCS8450 (Proxy)	1.321 ms	0 - 96 MB	NPU
ResNeXt50	TFLITE	w8a8	Qualcomm® SA7255P	2.223 ms	0 - 67 MB	NPU
ResNeXt50	TFLITE	w8a8	Qualcomm® SA8295P	1.586 ms	0 - 62 MB	NPU
ResNeXt50	TFLITE	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	0.551 ms	0 - 61 MB	NPU
ResNeXt50	TFLITE	w8a8	Snapdragon® 7 Gen 4 Mobile	1.279 ms	0 - 67 MB	NPU

License

• The license for the original implementation of ResNeXt50 can be found here.

References

• Aggregated Residual Transformations for Deep Neural Networks
• 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.