eNPU Predict

Predict responses of eNPU using compiled eAI model

Since R2025a

expand all in page

Libraries:
Embedded Coder Support Package for Qualcomm Hexagon Processors / Hexagon / eNPU Library

Description

The eNPU Predict block predicts responses by using a compiled embedded Artificial Intelligence (eAI ) model designed for Qualcomm® eNPU, based on the given input data. The block helps you to simulate and evaluate the neural network represented in the eAI network file.

Use this block to select a deep learning network represented as an eAI model (eAI network file) that is generated using the LPAI SDK Addon for Qualcomm Hexagon® SDK. For host simulations, the eNPU Predict block uses the LPAI SDK's host runtime library in the background to process the input data.

If the input data is of data type single, use the input value range option available in the block to quantize the input as per the selected eAI model.

The output data is quantized by default. To obtain single-precision output, select the dequantize option in the block and provide the output value range as per the selected eAI model.

Before predicting the response using the eAI model, you can also use the additional performance options available in the block to customize some of the performance configuration parameters of the eAI client. For more information on these parameters, refer to the Qualcomm LPAI SDK documentation.

Note

The eNPU Predict block does not support the Rapid Accelerator simulation mode. When you add the eNPU Predict block to your model, use the Normal or Accelerator modes.

Examples

Deploy Smart Speaker Model on Qualcomm Hexagon eNPU using LPAI SDK Add-on

Deploy Smart Speaker Model on Qualcomm Hexagon eNPU using LPAI SDK Add-on

Deploy a Simulink® model designed as smart speaker system on Qualcomm® Hexagon® eNPU using Embedded Coder® Support Package for Qualcomm Hexagon Processors. This example is based on the Simulink model available in Apply Speech Command Recognition Network in Smart Speaker Simulink Model example. In the updated model this example, the eNPU Predict block replaces the DL predict block from the original example as this block uses a precompiled eAI model for deploying the DL network to Hexagon eNPU.

Convert MATLAB Deep Learning Networks to eAI Model Using Qualcomm LPAI SDK

Convert MATLAB Deep Learning Networks to eAI Model Using Qualcomm LPAI SDK

Convert Deep Learning networks developed using MATLAB® built-in layers into eAI models for deployment to eNPU using Qualcomm® LPAI SDK. Deploying to an eNPU (LPAI) requires the Deep Learning network to be in the .eai file format (eAI model), generated using eai_builder tool from the LPAI SDK. This SDK provides a workflow for converting a ONNX/TFLite models to this format. The workflow explained in this example helps you to streamline the process for conversion of MATLAB-based Deep Learning networks to the .eai format.

Ports

Input

expand all

The input data, specified as an N-dimensional array. The array must be of the same size specified in the Input layer size parameter of the eAI model.

The eNPU Predict block supports multiple-input multiple-output tensor with a maximum of four dimensions, but the batch size must always be 1. For example, if the input layer of the original deep learning network is 128-by-128-by-3, the input signal dimension must be either 128-by-128-by-3 or 1-by-128-by-128-by-3.

If the leading dimensions are 1 (singleton dimensions), you can remove these dimensions without affecting compatibility. For example, if the input layer of an AI model expects an input of size 1-by-1-by-128-by-3, you can specify an input of size 1-by-1-by-128-by-3 or 128-by-3. You can remove these dimensions because dimensions of size 1 can be broadcast to match the expected shape.

The eNPU Predict block accepts either a floating-point (which requires quantization using other block parameters) or a fixed-point input signal.

Data type of input signal must match with the eAI model (eAI network file) if its input layer is of type int8, int16 or int32. If the input signal data type is single, specify the Input value range [Min Value, Max Value] parameter.

Data Types: single | int8 | int16 | int32

Output

expand all

The output tensor specified as an N-dimensional array. The array must be of the same size specified in the Output layer size parameter of the eAI model.

The eNPU Predict block supports a multiple-input multiple-output tensor with a maximum of four dimensions, but the batch size must always be 1.

Data type of output signal must match with the eAI model (eAI network file) if its input layer is of type int8, int16 or int32. If the output signal data type is single, select the Dequantize output parameter and specify the Output value range [MinValue,MaxValue] parameter.

Data Types: single | int8 | int16 | int32

Parameters

expand all

Click Browse and select the compiled eAI model (eAI network file) to predict responses. For more information on creating an eAI model using the eAI Builder tool, refer to the Qualcomm LPAI SDK documentation.

Minimum and maximum range of input values for quantization of floating-point input, specified as an M-by-2 array containing M number of [min,max] values, where M is the count of input layers.

For example, to quantize both inputs of a tensor layout with two inputs, specify the value: [[0, 255];[0,255]].

The block supports both symmetric quantization (absolute of minimum and maximum values are same) and asymmetric quantization (absolute of minimum and maximum values are different).

Select this parameter to dequantize the block's output.

Minimum and maximum range of output values for quantization of floating-point output, specified as an M-by-2 array containing M number of [min,max] values, where M is the count of output layers.

The block supports both symmetric quantization (absolute value of minimum and maximum values are same) and asymmetric quantization (absolute value of minimum and maximum values are different).

Dependencies

This parameter appears only if you select Dequantize output checkbox.

Select this parameter to use the default performance configuration values for the eAI client while obtaining response. The default values as per the LPAI SDK Addon documentation are:

  • Frames per second: 1

  • Faster Than Real Time (FTRT) ratio: 10

  • Priority level: 3

Number of frames to process per second while predicting response using the eAI model. For more information, refer to the Qualcomm LPAI SDK documentation.

Dependencies

To enable this parameter, clear the Use default performance configuration parameter.

Ratio to use for faster than real-time performance while predicting response using the eAI model. For more information, refer to the Qualcomm LPAI SDK documentation.

Dependencies

To enable this parameter, clear the Use default performance configuration parameter.

Client priority level for eNPU scheduling while predicting response using the eAI model. You can specify one of these values with the corresponding priorities:

  • 3 – Very low

  • 2 – Low

  • 1 – Medium

  • 0 – High

For more information, refer to the Qualcomm LPAI SDK documentation.

Dependencies

To enable this parameter, clear the Use default performance configuration parameter.

Select this parameter to perform real-time processing of input data while predicting the response.

Dependencies

To enable this parameter, clear the Use default performance configuration parameter.

Extended Capabilities

expand all

C/C++ Code Generation
Generate C and C++ code using Simulink® Coder™.

Version History

Introduced in R2025a