R-Car H3 SDK

This guide was created for versions: v0.1.0 - Latest

  • Purchase An R-Car H3 Starter Kit

    Purchase the R-Car H3 kit directly from approved Renesas distributors to access the OpenCL implementation.

  • Download the ComputeAorta Runtime

    ComputeAorta brings pre-conformant OpenCL support to R-Car H3 hardware. You can download the release package from our website using the button below.

  • Get Started with ComputeAorta for R-Car

    We have created a Getting Started guide for using OpenCL with the R-Car H3 hardware that explains the steps to setup and use ComputeAorta.

R-Car Starter Kit With OpenCL and SYCL

Software development for automotive applications including Advanced Driver Assistance Systems is increasing in scale and complexity. Developers need to be able to rapidly prototype and build applications for automotive systems using the latest technologies and software frameworks. By using ComputeAorta and ComputeCpp with Renesas H3 hardware, developers can write applications using the open standards OpenCL and SYCL. This enables the execution of applications on heterogeneous hardware providing acceleration and power efficiency.

Support for OpenCL will enable developers to build accelerated applications using the OpenCV framework, one of the most advanced vision frameworks available. TensorFlow offers developers the ability to write complex machine learning applications, and with ComputeCpp developers will be able to run these applications on Renesas H3 hardware.

By using open standards, developers can build applications using familiar C++ code that runs on current and future Renesas hardware, and can take advantage of the optimization and built in parallel capabilities of OpenCL and SYCL.

Codeplay and R-Car offers:

  1. Simplified development. ComputeAorta and ComputeCpp are designed to simplify the development process providing familiar programming environments, enabling developers to focus on writing complex applications.
  2. Parallel capabilities for developers. Both OpenCL and SYCL offer developers built in capabilities to run their applications in parallel on heterogeneous hardware. These abstractions bring more portability for developers, reducing the effort to develop applications to take advantage of parallel computing.
  3. Support for high-level computer vision frameworks. By offering OpenCL support it is now possible for developers to use OpenCV and other computer vision frameworks with R-Car H3 hardware, taking advantage of their parallel capabilities. TensorFlow has become one of the most popular machine learning frameworks and using ComputeCpp, our implementation of the SYCL standard, developers can run their TensorFlow applications on R-Car H3 hardware.