MNASNet05: Optimized for Qualcomm Devices

MNASNet05 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 MNASNet05 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 1.25.0	Download
ONNX	w8a16	Universal	QAIRT 2.45, 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 MNASNet05 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 MNASNet05 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: 2.21M
• Model size (float): 8.45 MB
• Model size (w8a16): 2.79 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
MNASNet05	ONNX	float	Snapdragon® X2 Elite	0.245 ms	2 - 2 MB	NPU
MNASNet05	ONNX	float	Snapdragon® X Elite	0.483 ms	5 - 5 MB	NPU
MNASNet05	ONNX	float	Snapdragon® 8 Gen 3 Mobile	0.331 ms	1 - 48 MB	NPU
MNASNet05	ONNX	float	Snapdragon® 8 Gen 1 Mobile	0.871 ms	1 - 51 MB	NPU
MNASNet05	ONNX	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	0.483 ms	0 - 7 MB	NPU
MNASNet05	ONNX	float	Qualcomm® QCS8450	0.871 ms	1 - 51 MB	NPU
MNASNet05	ONNX	float	Qualcomm® Dragonwing™ IQ-9075	0.812 ms	1 - 3 MB	NPU
MNASNet05	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	0.224 ms	0 - 28 MB	NPU
MNASNet05	ONNX	float	Snapdragon® 8 Elite Mobile	0.269 ms	0 - 32 MB	NPU
MNASNet05	ONNX	float	Qualcomm® Dragonwing™ Q-8750	0.269 ms	0 - 32 MB	NPU
MNASNet05	ONNX	float	Qualcomm® Dragonwing™ IQ-X7181	0.483 ms	5 - 5 MB	NPU
MNASNet05	ONNX	w8a16	Snapdragon® X2 Elite	0.215 ms	1 - 1 MB	NPU
MNASNet05	ONNX	w8a16	Snapdragon® X Elite	0.52 ms	0 - 0 MB	NPU
MNASNet05	ONNX	w8a16	Snapdragon® 8 Gen 3 Mobile	0.35 ms	0 - 44 MB	NPU
MNASNet05	ONNX	w8a16	Snapdragon® 8 Gen 1 Mobile	0.664 ms	0 - 48 MB	NPU
MNASNet05	ONNX	w8a16	Qualcomm® Dragonwing™ QCS6490	2.151 ms	0 - 3 MB	NPU
MNASNet05	ONNX	w8a16	Qualcomm® Dragonwing™ QCS8550 (Proxy)	0.518 ms	0 - 16 MB	NPU
MNASNet05	ONNX	w8a16	Qualcomm® QCS8450	0.664 ms	0 - 48 MB	NPU
MNASNet05	ONNX	w8a16	Qualcomm® Dragonwing™ IQ-9075	0.657 ms	0 - 3 MB	NPU
MNASNet05	ONNX	w8a16	Snapdragon® 7 Gen 4 Mobile	0.53 ms	0 - 31 MB	NPU
MNASNet05	ONNX	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	0.216 ms	0 - 33 MB	NPU
MNASNet05	ONNX	w8a16	Qualcomm® Dragonwing™ Q-6690	2.232 ms	0 - 145 MB	NPU
MNASNet05	ONNX	w8a16	Snapdragon® 8 Elite Mobile	0.259 ms	0 - 30 MB	NPU
MNASNet05	ONNX	w8a16	Qualcomm® Dragonwing™ Q-7790	0.53 ms	0 - 31 MB	NPU
MNASNet05	ONNX	w8a16	Qualcomm® Dragonwing™ Q-8750	0.259 ms	0 - 30 MB	NPU
MNASNet05	ONNX	w8a16	Qualcomm® Dragonwing™ IQ-X7181	0.52 ms	0 - 0 MB	NPU
MNASNet05	QNN_DLC	float	Snapdragon® X2 Elite	0.411 ms	1 - 1 MB	NPU
MNASNet05	QNN_DLC	float	Snapdragon® X Elite	0.915 ms	1 - 1 MB	NPU
MNASNet05	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	0.509 ms	0 - 44 MB	NPU
MNASNet05	QNN_DLC	float	Snapdragon® 8 Gen 1 Mobile	1.564 ms	0 - 47 MB	NPU
MNASNet05	QNN_DLC	float	Qualcomm® QCS8275	2.347 ms	1 - 28 MB	NPU
MNASNet05	QNN_DLC	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	0.783 ms	1 - 16 MB	NPU
MNASNet05	QNN_DLC	float	Qualcomm® SA8775P	1.102 ms	0 - 30 MB	NPU
MNASNet05	QNN_DLC	float	Qualcomm® SA8650P	1.102 ms	0 - 30 MB	NPU
MNASNet05	QNN_DLC	float	Qualcomm® SA8255P	1.102 ms	0 - 30 MB	NPU
MNASNet05	QNN_DLC	float	Qualcomm® QCS8450	1.564 ms	0 - 47 MB	NPU
MNASNet05	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-9075	1.102 ms	1 - 3 MB	NPU
MNASNet05	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	0.295 ms	1 - 29 MB	NPU
MNASNet05	QNN_DLC	float	Qualcomm® SA7255P	2.347 ms	1 - 28 MB	NPU
MNASNet05	QNN_DLC	float	Snapdragon® 8 Elite Mobile	0.389 ms	0 - 28 MB	NPU
MNASNet05	QNN_DLC	float	Qualcomm® SA8295P	1.437 ms	0 - 28 MB	NPU
MNASNet05	QNN_DLC	float	Qualcomm® Dragonwing™ Q-8750	0.389 ms	0 - 28 MB	NPU
MNASNet05	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-X7181	0.915 ms	1 - 1 MB	NPU
MNASNet05	QNN_DLC	w8a16	Snapdragon® X2 Elite	0.407 ms	0 - 0 MB	NPU
MNASNet05	QNN_DLC	w8a16	Snapdragon® X Elite	0.891 ms	0 - 0 MB	NPU
MNASNet05	QNN_DLC	w8a16	Snapdragon® 8 Gen 3 Mobile	0.528 ms	0 - 39 MB	NPU
MNASNet05	QNN_DLC	w8a16	Snapdragon® 8 Gen 1 Mobile	0.953 ms	0 - 45 MB	NPU
MNASNet05	QNN_DLC	w8a16	Qualcomm® Dragonwing™ QCS6490	2.72 ms	0 - 2 MB	NPU
MNASNet05	QNN_DLC	w8a16	Qualcomm® QCS8275	1.656 ms	0 - 27 MB	NPU
MNASNet05	QNN_DLC	w8a16	Qualcomm® Dragonwing™ QCS8550 (Proxy)	0.767 ms	0 - 13 MB	NPU
MNASNet05	QNN_DLC	w8a16	Qualcomm® SA8775P	0.955 ms	0 - 29 MB	NPU
MNASNet05	QNN_DLC	w8a16	Qualcomm® SA8650P	0.955 ms	0 - 29 MB	NPU
MNASNet05	QNN_DLC	w8a16	Qualcomm® SA8255P	0.955 ms	0 - 29 MB	NPU
MNASNet05	QNN_DLC	w8a16	Qualcomm® QCS8450	0.953 ms	0 - 45 MB	NPU
MNASNet05	QNN_DLC	w8a16	Qualcomm® Dragonwing™ IQ-9075	0.859 ms	2 - 4 MB	NPU
MNASNet05	QNN_DLC	w8a16	Snapdragon® 7 Gen 4 Mobile	0.79 ms	0 - 27 MB	NPU
MNASNet05	QNN_DLC	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	0.298 ms	0 - 29 MB	NPU
MNASNet05	QNN_DLC	w8a16	Qualcomm® Dragonwing™ Q-6690	3.056 ms	0 - 143 MB	NPU
MNASNet05	QNN_DLC	w8a16	Qualcomm® SA7255P	1.656 ms	0 - 27 MB	NPU
MNASNet05	QNN_DLC	w8a16	Snapdragon® 8 Elite Mobile	0.351 ms	0 - 28 MB	NPU
MNASNet05	QNN_DLC	w8a16	Qualcomm® SA8295P	1.232 ms	0 - 25 MB	NPU
MNASNet05	QNN_DLC	w8a16	Qualcomm® Dragonwing™ Q-7790	0.79 ms	0 - 27 MB	NPU
MNASNet05	QNN_DLC	w8a16	Qualcomm® Dragonwing™ Q-8750	0.351 ms	0 - 28 MB	NPU
MNASNet05	QNN_DLC	w8a16	Qualcomm® Dragonwing™ IQ-X7181	0.891 ms	0 - 0 MB	NPU
MNASNet05	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	0.524 ms	0 - 46 MB	NPU
MNASNet05	TFLITE	float	Snapdragon® 8 Gen 1 Mobile	1.579 ms	0 - 50 MB	NPU
MNASNet05	TFLITE	float	Qualcomm® QCS8275	2.359 ms	0 - 28 MB	NPU
MNASNet05	TFLITE	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	0.796 ms	0 - 2 MB	NPU
MNASNet05	TFLITE	float	Qualcomm® SA8775P	1.109 ms	0 - 31 MB	NPU
MNASNet05	TFLITE	float	Qualcomm® SA8650P	1.109 ms	0 - 31 MB	NPU
MNASNet05	TFLITE	float	Qualcomm® SA8255P	1.109 ms	0 - 31 MB	NPU
MNASNet05	TFLITE	float	Qualcomm® QCS8450	1.579 ms	0 - 50 MB	NPU
MNASNet05	TFLITE	float	Qualcomm® Dragonwing™ IQ-9075	0.982 ms	0 - 8 MB	NPU
MNASNet05	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	0.295 ms	0 - 30 MB	NPU
MNASNet05	TFLITE	float	Qualcomm® SA7255P	2.359 ms	0 - 28 MB	NPU
MNASNet05	TFLITE	float	Snapdragon® 8 Elite Mobile	0.382 ms	0 - 28 MB	NPU
MNASNet05	TFLITE	float	Qualcomm® SA8295P	1.441 ms	0 - 28 MB	NPU
MNASNet05	TFLITE	float	Qualcomm® Dragonwing™ Q-8750	0.382 ms	0 - 28 MB	NPU

License

• The license for the original implementation of MNASNet05 can be found here.

References

• MnasNet: Platform-Aware Neural Architecture Search for Mobile
• 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.