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 2.45, ONNX Runtime 1.27.1	Download
ONNX	w8a16	Universal	QAIRT 2.45, ONNX Runtime 1.27.1	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® X2 Elite	3.047 ms	2 - 2 MB	NPU
EfficientNet-V2-s	ONNX	float	Snapdragon® X Elite	5.603 ms	46 - 46 MB	NPU
EfficientNet-V2-s	ONNX	float	Snapdragon® 8 Gen 3 Mobile	4.048 ms	0 - 165 MB	NPU
EfficientNet-V2-s	ONNX	float	Snapdragon® 8 Gen 1 Mobile	11.615 ms	0 - 194 MB	NPU
EfficientNet-V2-s	ONNX	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	5.335 ms	0 - 222 MB	NPU
EfficientNet-V2-s	ONNX	float	Qualcomm® QCS8450	11.615 ms	0 - 194 MB	NPU
EfficientNet-V2-s	ONNX	float	Qualcomm® Dragonwing™ IQ-9075	7.632 ms	1 - 6 MB	NPU
EfficientNet-V2-s	ONNX	float	Qualcomm® Dragonwing™ IQ-X7181	5.603 ms	46 - 46 MB	NPU
EfficientNet-V2-s	ONNX	float	Qualcomm® Dragonwing™ Q-8750	3.095 ms	0 - 199 MB	NPU
EfficientNet-V2-s	ONNX	float	Snapdragon® 8 Elite Mobile	3.095 ms	0 - 199 MB	NPU
EfficientNet-V2-s	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	2.322 ms	0 - 204 MB	NPU
EfficientNet-V2-s	ONNX	w8a16	Snapdragon® X2 Elite	2.368 ms	2 - 2 MB	NPU
EfficientNet-V2-s	ONNX	w8a16	Snapdragon® X Elite	5.705 ms	24 - 24 MB	NPU
EfficientNet-V2-s	ONNX	w8a16	Snapdragon® 8 Gen 3 Mobile	3.64 ms	0 - 210 MB	NPU
EfficientNet-V2-s	ONNX	w8a16	Snapdragon® 8 Gen 1 Mobile	6.85 ms	1 - 218 MB	NPU
EfficientNet-V2-s	ONNX	w8a16	Qualcomm® Dragonwing™ QCS6490	20.177 ms	1 - 4 MB	NPU
EfficientNet-V2-s	ONNX	w8a16	Qualcomm® Dragonwing™ QCS8550 (Proxy)	5.46 ms	0 - 30 MB	NPU
EfficientNet-V2-s	ONNX	w8a16	Qualcomm® QCS8450	6.85 ms	1 - 218 MB	NPU
EfficientNet-V2-s	ONNX	w8a16	Qualcomm® Dragonwing™ IQ-9075	5.749 ms	1 - 4 MB	NPU
EfficientNet-V2-s	ONNX	w8a16	Qualcomm® Dragonwing™ IQ-X7181	5.705 ms	24 - 24 MB	NPU
EfficientNet-V2-s	ONNX	w8a16	Qualcomm® Dragonwing™ Q-8750	2.52 ms	0 - 164 MB	NPU
EfficientNet-V2-s	ONNX	w8a16	Snapdragon® 8 Elite Mobile	2.52 ms	0 - 164 MB	NPU
EfficientNet-V2-s	ONNX	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	2.023 ms	0 - 167 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Snapdragon® X2 Elite	3.492 ms	2 - 2 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Snapdragon® X Elite	6.279 ms	2 - 2 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	4.336 ms	2 - 162 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Snapdragon® 8 Gen 1 Mobile	13.477 ms	2 - 192 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Qualcomm® Dragonwing™ QCS8275	25.722 ms	2 - 75 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	5.785 ms	0 - 9 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Qualcomm® SA8775P	8.22 ms	2 - 77 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Qualcomm® SA8650P	8.22 ms	2 - 77 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Qualcomm® SA8255P	8.22 ms	2 - 77 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Qualcomm® QCS8450	13.477 ms	2 - 192 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-9075	7.907 ms	2 - 5 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-X7181	6.279 ms	2 - 2 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Qualcomm® Dragonwing™ Q-8750	3.133 ms	0 - 79 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Qualcomm® SA7255P	25.722 ms	2 - 75 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Qualcomm® SA8295P	13.006 ms	2 - 107 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Snapdragon® 8 Elite Mobile	3.133 ms	0 - 79 MB	NPU
EfficientNet-V2-s	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	2.285 ms	2 - 85 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Snapdragon® X2 Elite	2.91 ms	1 - 1 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Snapdragon® X Elite	6.83 ms	1 - 1 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Snapdragon® 8 Gen 3 Mobile	4.201 ms	1 - 175 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Snapdragon® 8 Gen 1 Mobile	8.212 ms	1 - 185 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Qualcomm® Dragonwing™ QCS6490	21.506 ms	3 - 5 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Qualcomm® Dragonwing™ QCS8275	12.065 ms	1 - 134 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Qualcomm® Dragonwing™ QCS8550 (Proxy)	6.233 ms	1 - 3 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Qualcomm® SA8775P	6.716 ms	1 - 136 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Qualcomm® SA8650P	6.716 ms	1 - 136 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Qualcomm® SA8255P	6.716 ms	1 - 136 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Qualcomm® QCS8450	8.212 ms	1 - 185 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Qualcomm® Dragonwing™ IQ-9075	6.464 ms	1 - 3 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Qualcomm® Dragonwing™ IQ-X7181	6.83 ms	1 - 1 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Qualcomm® Dragonwing™ Q-6690	41.3 ms	1 - 255 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Qualcomm® Dragonwing™ Q-7790	6.973 ms	1 - 255 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Qualcomm® Dragonwing™ Q-8750	2.809 ms	1 - 146 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Qualcomm® SA7255P	12.065 ms	1 - 134 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Qualcomm® SA8295P	8.384 ms	1 - 134 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Snapdragon® 8 Elite Mobile	2.809 ms	1 - 146 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	2.243 ms	1 - 145 MB	NPU
EfficientNet-V2-s	QNN_DLC	w8a16	Snapdragon® 7 Gen 4 Mobile	6.973 ms	1 - 255 MB	NPU
EfficientNet-V2-s	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	4.323 ms	0 - 200 MB	NPU
EfficientNet-V2-s	TFLITE	float	Snapdragon® 8 Gen 1 Mobile	13.412 ms	0 - 229 MB	NPU
EfficientNet-V2-s	TFLITE	float	Qualcomm® Dragonwing™ QCS8275	25.819 ms	0 - 116 MB	NPU
EfficientNet-V2-s	TFLITE	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	5.732 ms	0 - 51 MB	NPU
EfficientNet-V2-s	TFLITE	float	Qualcomm® SA8775P	8.218 ms	0 - 118 MB	NPU
EfficientNet-V2-s	TFLITE	float	Qualcomm® SA8650P	8.218 ms	0 - 118 MB	NPU
EfficientNet-V2-s	TFLITE	float	Qualcomm® SA8255P	8.218 ms	0 - 118 MB	NPU
EfficientNet-V2-s	TFLITE	float	Qualcomm® QCS8450	13.412 ms	0 - 229 MB	NPU
EfficientNet-V2-s	TFLITE	float	Qualcomm® Dragonwing™ IQ-9075	7.918 ms	0 - 51 MB	NPU
EfficientNet-V2-s	TFLITE	float	Qualcomm® Dragonwing™ Q-8750	3.145 ms	0 - 122 MB	NPU
EfficientNet-V2-s	TFLITE	float	Qualcomm® SA7255P	25.819 ms	0 - 116 MB	NPU
EfficientNet-V2-s	TFLITE	float	Qualcomm® SA8295P	13.025 ms	0 - 148 MB	NPU
EfficientNet-V2-s	TFLITE	float	Snapdragon® 8 Elite Mobile	3.145 ms	0 - 122 MB	NPU
EfficientNet-V2-s	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	2.338 ms	0 - 120 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.