EfficientNet-V2-s: Optimized for Qualcomm Devices

EfficientNetV2-s 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 EfficientNet-V2-s 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 1.25.0	Download
ONNX	w8a16	Universal	QAIRT 2.42, ONNX Runtime 1.25.0	Download
QNN_DLC	float	Universal	QAIRT 2.45	Download
QNN_DLC	w8a16	Universal	QAIRT 2.45	Download
TFLITE	float	Universal	QAIRT 2.45	Download

For more device-specific assets and performance metrics, visit EfficientNet-V2-s 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 EfficientNet-V2-s on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.image_classification

Model Stats:

• Model checkpoint: Imagenet
• Input resolution: 384x384
• Number of parameters: 21.4M
• Model size (float): 81.7 MB
• Model size (w8a16): 27.2 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
EfficientNet-V2-s	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	2.237 ms	1 - 206 MB	NPU
EfficientNet-V2-s	ONNX	float	Snapdragon® X2 Elite	3.056 ms	210 - 210 MB	NPU
EfficientNet-V2-s	ONNX	float	Snapdragon® X Elite	5.579 ms	147 - 147 MB	NPU
EfficientNet-V2-s	ONNX	float	Snapdragon® 8 Gen 3 Mobile	4.067 ms	0 - 173 MB	NPU
EfficientNet-V2-s	ONNX	float	Qualcomm® QCS8550 (Proxy)	5.381 ms	2 - 156 MB	NPU
EfficientNet-V2-s	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	3.068 ms	0 - 204 MB	NPU
EfficientNet-V2-s	ONNX	float	Qualcomm® QCS9075	8.057 ms	2 - 47 MB	NPU
EfficientNet-V2-s	ONNX	float	Qualcomm® QCS8750	3.068 ms	0 - 204 MB	NPU
EfficientNet-V2-s	ONNX	float	Qualcomm® QCS7181	5.579 ms	147 - 147 MB	NPU
EfficientNet-V2-s	ONNX	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	1.974 ms	0 - 159 MB	NPU
EfficientNet-V2-s	ONNX	w8a16	Snapdragon® X2 Elite	2.318 ms	211 - 211 MB	NPU
EfficientNet-V2-s	ONNX	w8a16	Snapdragon® X Elite	5.683 ms	180 - 180 MB	NPU
EfficientNet-V2-s	ONNX	w8a16	Snapdragon® 8 Gen 3 Mobile	3.665 ms	0 - 220 MB	NPU
EfficientNet-V2-s	ONNX	w8a16	Qualcomm® QCS6490	654.38 ms	60 - 64 MB	CPU
EfficientNet-V2-s	ONNX	w8a16	Qualcomm® QCS8550 (Proxy)	5.481 ms	0 - 48 MB	NPU
EfficientNet-V2-s	ONNX	w8a16	Qualcomm® QCM6690	327.766 ms	50 - 63 MB	CPU
EfficientNet-V2-s	ONNX	w8a16	Snapdragon® 7 Gen 4 Mobile	299.933 ms	57 - 71 MB	CPU
EfficientNet-V2-s	ONNX	w8a16	Snapdragon® 8 Elite For Galaxy Mobile	2.512 ms	0 - 159 MB	NPU
EfficientNet-V2-s	ONNX	w8a16	Qualcomm® QCS9075	5.988 ms	1 - 46 MB	NPU
EfficientNet-V2-s	ONNX	w8a16	Qualcomm® QCS7790	299.933 ms	57 - 71 MB	CPU
EfficientNet-V2-s	ONNX	w8a16	Qualcomm® QCS8750	2.512 ms	0 - 159 MB	NPU
EfficientNet-V2-s	ONNX	w8a16	Qualcomm® QCS7181	5.683 ms	180 - 180 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	2.273 ms	2 - 84 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Snapdragon® X2 Elite	3.439 ms	2 - 2 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Snapdragon® X Elite	6.392 ms	2 - 2 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	4.333 ms	0 - 157 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	5.777 ms	2 - 3 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Qualcomm® SA8775P	8.199 ms	2 - 77 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Qualcomm® SA8650P	8.199 ms	2 - 77 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Qualcomm® SA8255P	8.199 ms	2 - 77 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	13.365 ms	0 - 188 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Qualcomm® SA8295P	13.001 ms	2 - 108 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	3.134 ms	0 - 78 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Qualcomm® QCS9075	8.752 ms	4 - 7 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Qualcomm® QCS8750	3.134 ms	0 - 78 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Qualcomm® QCS7181	6.392 ms	2 - 2 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	2.202 ms	1 - 146 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Snapdragon® X2 Elite	2.899 ms	1 - 1 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Snapdragon® X Elite	6.825 ms	1 - 1 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Snapdragon® 8 Gen 3 Mobile	4.209 ms	0 - 175 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Qualcomm® QCS6490	18.735 ms	3 - 5 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Qualcomm® QCS8275	12.084 ms	1 - 134 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Qualcomm® QCS8550 (Proxy)	6.268 ms	1 - 162 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Qualcomm® SA8775P	6.734 ms	1 - 137 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Qualcomm® SA8650P	6.734 ms	1 - 137 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Qualcomm® SA8255P	6.734 ms	1 - 137 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Qualcomm® QCM6690	40.373 ms	1 - 253 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Qualcomm® SA7255P	12.084 ms	1 - 134 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Snapdragon® 7 Gen 4 Mobile	7.213 ms	1 - 254 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Qualcomm® SA8295P	8.366 ms	1 - 135 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Snapdragon® 8 Elite For Galaxy Mobile	2.811 ms	1 - 146 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Qualcomm® QCS9075	6.911 ms	1 - 3 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Qualcomm® QCS8450 (Proxy)	8.085 ms	1 - 185 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Qualcomm® QCS7790	7.213 ms	1 - 254 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Qualcomm® QCS8750	2.811 ms	1 - 146 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Qualcomm® QCS7181	6.825 ms	1 - 1 MB	NPU
EfficientNet-V2-s	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	2.251 ms	0 - 124 MB	NPU
EfficientNet-V2-s	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	4.333 ms	0 - 199 MB	NPU
EfficientNet-V2-s	TFLITE	float	Qualcomm® QCS8275	25.837 ms	0 - 116 MB	NPU
EfficientNet-V2-s	TFLITE	float	Qualcomm® QCS8550 (Proxy)	5.735 ms	0 - 2 MB	NPU
EfficientNet-V2-s	TFLITE	float	Qualcomm® SA8775P	8.228 ms	0 - 118 MB	NPU
EfficientNet-V2-s	TFLITE	float	Qualcomm® SA8650P	8.228 ms	0 - 118 MB	NPU
EfficientNet-V2-s	TFLITE	float	Qualcomm® SA8255P	8.228 ms	0 - 118 MB	NPU
EfficientNet-V2-s	TFLITE	float	Qualcomm® QCS8450 (Proxy)	13.441 ms	0 - 229 MB	NPU
EfficientNet-V2-s	TFLITE	float	Qualcomm® SA8295P	13.037 ms	0 - 148 MB	NPU
EfficientNet-V2-s	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	3.137 ms	0 - 123 MB	NPU
EfficientNet-V2-s	TFLITE	float	Qualcomm® QCS9075	8.765 ms	0 - 51 MB	NPU
EfficientNet-V2-s	TFLITE	float	Qualcomm® SA7255P	25.837 ms	0 - 116 MB	NPU
EfficientNet-V2-s	TFLITE	float	Qualcomm® QCS8750	3.137 ms	0 - 123 MB	NPU

License

• The license for the original implementation of EfficientNet-V2-s can be found here.

References

• EfficientNetV2: Smaller Models and Faster Training
• 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.