You can deploy this model and start earning money today!

qualcomm/
NASNet

Views

No views yet

NASNet: Optimized for Qualcomm Devices

NASNet is a CNN-based architecture discovered via Neural Architecture Search (NAS) that can classify images from the Imagenet dataset.

This is based on the implementation of NASNet 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
QNN_DLC	float	Universal	QAIRT 2.45	Download
TFLITE	float	Universal	QAIRT 2.45	Download

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

Model Details

Model Type: Model_use_case.image_classification

Model Stats:

Activations (M): 19.4
GMACs: 5.9
Input resolution: 224x224
Model checkpoint: nasnetalarge.tf_in1k
Model size (float): 338 MB
Number of parameters: 88.7M

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
NASNet	ONNX	float	Snapdragon® X2 Elite	9.642 ms	1 - 1 MB	NPU
NASNet	ONNX	float	Snapdragon® X Elite	18.866 ms	188 - 188 MB	NPU
NASNet	ONNX	float	Snapdragon® 8 Gen 3 Mobile	13.854 ms	0 - 572 MB	NPU
NASNet	ONNX	float	Snapdragon® 8 Gen 1 Mobile	53.3 ms	2 - 521 MB	NPU
NASNet	ONNX	float	Qualcomm® Dragonwing™ IQ-8275	30.091 ms	1 - 6 MB	NPU
NASNet	ONNX	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	18.883 ms	0 - 193 MB	NPU
NASNet	ONNX	float	Qualcomm® QCS8450	53.3 ms	2 - 521 MB	NPU
NASNet	ONNX	float	Qualcomm® Dragonwing™ IQ-9075	27.932 ms	1 - 5 MB	NPU
NASNet	ONNX	float	Qualcomm® Dragonwing™ IQ-X7181	18.866 ms	188 - 188 MB	NPU
NASNet	ONNX	float	Qualcomm® Dragonwing™ Q-8750	11.695 ms	1 - 406 MB	NPU
NASNet	ONNX	float	Snapdragon® 8 Elite Mobile	11.695 ms	1 - 406 MB	NPU
NASNet	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	9.494 ms	1 - 419 MB	NPU
NASNet	QNN_DLC	float	Snapdragon® X2 Elite	8.931 ms	1 - 1 MB	NPU
NASNet	QNN_DLC	float	Snapdragon® X Elite	16.791 ms	1 - 1 MB	NPU
NASNet	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	12.173 ms	0 - 513 MB	NPU
NASNet	QNN_DLC	float	Snapdragon® 8 Gen 1 Mobile	48.598 ms	0 - 473 MB	NPU
NASNet	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-8275	27.588 ms	1 - 5 MB	NPU
NASNet	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-8275	89.183 ms	1 - 350 MB	NPU
NASNet	QNN_DLC	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	16.25 ms	1 - 3 MB	NPU
NASNet	QNN_DLC	float	Qualcomm® SA8775P	24.425 ms	1 - 355 MB	NPU
NASNet	QNN_DLC	float	Qualcomm® SA8650P	24.425 ms	1 - 355 MB	NPU
NASNet	QNN_DLC	float	Qualcomm® SA8255P	24.425 ms	1 - 355 MB	NPU
NASNet	QNN_DLC	float	Qualcomm® QCS8450	48.598 ms	0 - 473 MB	NPU
NASNet	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-9075	24.269 ms	1 - 4 MB	NPU
NASNet	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-X7181	16.791 ms	1 - 1 MB	NPU
NASNet	QNN_DLC	float	Qualcomm® Dragonwing™ Q-8750	9.773 ms	1 - 351 MB	NPU
NASNet	QNN_DLC	float	Qualcomm® SA7255P	89.183 ms	1 - 350 MB	NPU
NASNet	QNN_DLC	float	Qualcomm® SA8295P	38.608 ms	1 - 310 MB	NPU
NASNet	QNN_DLC	float	Snapdragon® 8 Elite Mobile	9.773 ms	1 - 351 MB	NPU
NASNet	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	7.812 ms	0 - 365 MB	NPU
NASNet	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	12.718 ms	0 - 695 MB	NPU
NASNet	TFLITE	float	Snapdragon® 8 Gen 1 Mobile	49.371 ms	0 - 653 MB	NPU
NASNet	TFLITE	float	Qualcomm® Dragonwing™ IQ-8275	28.044 ms	0 - 193 MB	NPU
NASNet	TFLITE	float	Qualcomm® Dragonwing™ IQ-8275	89.668 ms	0 - 537 MB	NPU
NASNet	TFLITE	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	16.773 ms	0 - 2 MB	NPU
NASNet	TFLITE	float	Qualcomm® SA8775P	24.835 ms	0 - 537 MB	NPU
NASNet	TFLITE	float	Qualcomm® SA8650P	24.835 ms	0 - 537 MB	NPU
NASNet	TFLITE	float	Qualcomm® SA8255P	24.835 ms	0 - 537 MB	NPU
NASNet	TFLITE	float	Qualcomm® QCS8450	49.371 ms	0 - 653 MB	NPU
NASNet	TFLITE	float	Qualcomm® Dragonwing™ IQ-9075	24.896 ms	0 - 192 MB	NPU
NASNet	TFLITE	float	Qualcomm® Dragonwing™ Q-8750	10.374 ms	0 - 533 MB	NPU
NASNet	TFLITE	float	Qualcomm® SA7255P	89.668 ms	0 - 537 MB	NPU
NASNet	TFLITE	float	Qualcomm® SA8295P	40.131 ms	0 - 487 MB	NPU
NASNet	TFLITE	float	Snapdragon® 8 Elite Mobile	10.374 ms	0 - 533 MB	NPU
NASNet	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	8.246 ms	0 - 548 MB	NPU

License

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

References

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.

qualcomm/NASNet