This is the procedure from setting up the evaluation environment to running the sample program. (Click here for RZ/V2L)

Please refer the AI Evaluation software for the evaluation of DRP-AI itself. Also, refer to the RZ/V2L, RZ/V2M, RZ/V2MA AI Evaluation Software Guide revision 7.40 (PDF | English, 日本語) in DRP-AI Support Package for information on how to use the evaluation software.

Learn more about the RZ/V Series

RZ/V2MA MPU

You can easily try out AI applications using RZ/V2MA MPUs by following the steps below.


Step 1: Obtain an evaluation board

The RZ/V2MA Evaluation Board Kit (EVK) is the ideal board kit for RZ/V2MA evaluation. You can start evaluating RZ/V2MA immediately by building an environment according to the following procedure.

Get RZ/V2MA Evaluation Board Kit

Image
RZV2MA Evaluation Board Kit

Step 2: Obtain Linux environment

Linux Package for RZ/V2MA is available. Please obtain the latest environment from

RZ/V Verified Linux Package v3.0.4 (ZIP)

Linux Package for RZ/V2MA is available. Please obtain the latest environment from

RZ/V Verified Linux PackageStart-Up Guidefor RZ/V2MA Rev.1.20 (PDF)
(This document is included in the RZ/V Verified Linux Package.)

<Points to consider when referring to the Start Up Guide>

  • For more information about VLP, please refer to RZ/V Verified Linux Package [5.10-CIP].
  • Please prepare a Linux PC environment for Table 1-1.
  • Follow the instructions in Chapter 2.Building Instructions to build. This may take several hours.
  • Follow the instructions in Chapter 2 Building Instructions (step4) when building in an offline environment.
  • Please download the Bootloader Package from Boot Loader Package for RZ/V2MA(ZIP).

Step 3: EVK setup

Using the prepared environment, let's start the Linux environment on EVK.

RZ/V Verified Linux PackageStart-Up Guidefor RZ/V2MA Rev.1.20 (PDF)
(This document is included in the RZ/V Verified Linux Package.)

Step3-1: Create an SD card for Flash writer on Ubuntu PC.

<point>

  • As described in Tabel3-1, use a micro-SDHC card, and only one partition formatted with FAT32.
  • Store the “B2_IinitSW.bin” generated in Step 2 to SD Card.
    (These files are in build/tmp/deploy/images/rzv2ma under your build execution directory. Use the files described in Table 2-2.)

Step3-2: Write Bootloader to EVK using Flash Writer.

<point>

  • Follow the steps in Chapter 4.2.2 to write Bootloader/U-Boot.
  • Use the Bootloader and U-Boot generated in Step 2. (These files are in build/tmp/deploy/images/rzv2ma under your build execution directory. Use the files described in Table 2-2.)
  • Although a boot loader and U-Boot are written to the EVK at the time of shipment, please be sure to write the boot loader and U-Boot generated in Step 2 to your EVK before use.
  • If you use network boot, please refer to chapter 4 of RZ/V2MA DRP-AI Support Package Version 7.40 Release Note (PDF)

Step3-3: Create an SD card on Ubuntu PC.

<point>

  • Create a partition formatted with FAT32 and ext4 on the SD card according to the procedure in chapter 3.1.2.
  • Write the Linux Kernel/Device tree/root filesystem created in Step 2.

Step3-4: Insert the created SD card into the EVK board and boot the Linux environment.

<point>

  • Follow the procedure in step5 of chapter 4.2.2 and chapter 5.1, 5.2 to change the DSW301 settings, set the U-Boot environment variable, and then start the Kernel.

Now the Linux environment has been started (Login).


Step 4: Apply DRP-AI environment

Next, the DRP-AI environment is applied. The RZ/V2MAL DRP-AI Support Package is available on Ubuntu PCs to support the DRP-AI functionality of RZ/V2MA. Please obtain the latest environment from

RZ/V2MA DRP-AI Support Package Version 7.40 (ZIP)

Please follow the instructions in the following document when setting up your environment.

RZ/V2MA DRP-AI Support Package Version 7.40 Release Note (PDF)
(This document is included in the RZ/V2MA DRP-AI Support Package.)

<point>

  • For more information about DRP-AI Support Package, please refer to RZ/V2L DRP-AI Support Package.
  • Build the environment created in Step 2 according to the procedures in Section 3.2.
  • Replace the data written to the SD card up to Step 3 with the data created in this step.

Step 5: Build the DRP-AI sample application

Build the sample application by creating an SDK environment in Ubuntu PC to cross-compile the sample application included in the RZ/V2MA DRP-AI Support Package obtained in Step 4.

Step5-1: Build the SDK environment.

RZ/V2MA DRP-AI Support Package Version 7.40 Release Note (PDF)
(This document is included in the RZ/V2MA DRP-AI Support Package.)

<point>

  • Follow the steps in chapter 6 to compile and configure the SDK environment.

Step5-2: Build the sample application in the SDK environment.

RZ/V2MA DRP-AI Sample Application Note revision 7.40 (PDF)
(This document is included in the RZ/V2MA DRP-AI Support Package.)

<point>

  • Follow the steps in Chapter 2 to build the sample application.
  • Set the working folder as "APP_WORK" and deploy the sample application there.
  • Table 1.2 in Sample Application Note lists the sample applications included in the package. For details of the sample applications, please refer to their respective pages.
  • Replace the Kernel Image, Device Tree and Root file System on the SD card with the data created in Step4.
  • We recommend that you first try the application (USB Camera HTTP version application). This application requires USB Camera specifications to YUYV format/VGA size/30fps.

Step5-3: Save the built application to SD card.

RZ/V2MA DRP-AI Sample Application Note revision 7.40 (PDF)
(This document is included in the RZ/V2MA DRP-AI Support Package.)

<point>

  • Save the application execution environment created to the SD card created in Step 4 according to the procedures in Chapter 3.

Step 6: Run the DRP-AI sample application

Insert the SD card created in Step5-3 into the EVK configured in Step4 and run the sample application.

Step6-1: Power on the board and boot Linux.

<point>

  • Since EVK has already been set up in Step 3, Linux will boot simply by turning on the power.

Step6-2: Run the application.

RZ/V2MA DRP-AI Sample Application Note revision 7.40 (PDF)
(This document is included in the RZ/V2MA DRP-AI Support Package.)

<point>

  • Follow the steps in Chapter 4 to run the sample application.
  • Please refer to Figure 3.1 for the required equipment.

Related Information:

RENESAS.info
(Information for engineers, from engineers)


Option1: Change the AI model

If you wish to use a model other than the AI model provided as a sample, please follow the guide below.

RZ/V2L, RZ/V2M, RZ/V2MA AI Implementation Guide Get Started revision 7.40 (PDF | English, 日本語)
(This document is included in the RZ/V2MA DRP-AI Support Package.)

<point>

  • The Implementation Guide is divided into a common part (Get Started) and an individual part (Pytorch ResNet, Darknet YOLO, etc.). The link above is for Get Started only. Please obtain the model-specific parts from the RZ/V2MA DRP-AI Support Package according to the model you are using.