Get Started with GPU Coder

GPU Coder™ generates optimized CUDA^® code from MATLAB^® code and Simulink^® models. The generated code includes CUDA kernels for parallelizable parts of your deep learning, embedded vision, and radar and signal processing algorithms. For high performance, the generated code can call NVIDIA^® TensorRT™. You can integrate the generated CUDA into your project as source code or static/dynamic libraries and compile it for modern NVIDIA GPUs, including those embedded on NVIDIA Jetson™ and NVIDIA DRIVE^® platforms. You can access peripherals on the Jetson and DRIVE platforms and incorporate manually written CUDA into the generated code.

GPU Coder enables you to profile the generated CUDA to identify bottlenecks and opportunities for performance optimization (with Embedded Coder^®). Bidirectional links let you trace between MATLAB code and generated CUDA. You can verify the numerical behavior of the generated code via software-in-the-loop (SIL) and processor-in-the-loop (PIL) testing.

Tutorials

MATLAB

• Code Generation by Using the GPU Coder App
  Generate CUDA code from MATLAB code by using the GPU Coder app.
• Code Generation Using the Command Line Interface
  Generate CUDA code from MATLAB code by using the codegen command.
• Verify Correctness of the Generated Code
  Behavioral verification of generated code, traceability, and code generation reports.
• Code Generation for Deep Learning Networks by Using cuDNN
  Generate code for pretrained convolutional neural networks by using the cuDNN library.
• Code Generation for Deep Learning Networks by Using TensorRT
  Generate code for pretrained convolutional neural networks by using the TensorRT library.
• Debug CUDA MEX Functions
  Suggestions for debugging CUDA MEX function.

Simulink

• Accelerate Simulation Speed by Using GPU Coder
  Achieve faster simulation for models that contain MATLAB Function blocks.
• Code Generation from Simulink Models with GPU Coder
  Generate CUDA code from Simulink models by using GPU Coder.
• GPU Code Generation for Blocks from the Deep Neural Networks Library
  Simulate and generate code for deep learning models in Simulink using library blocks.

About Code Generation from MATLAB Algorithms

• GPU Programming Paradigm
  

  Introduction to GPU accelerated computing.

• GPU Code Generation Workflow
  

  Design, implement, and verify generated CUDA MEX for acceleration and standalone CUDA code for deployment.

Featured Examples

GPU Code Generation: The Mandelbrot Set

Generate CUDA® code from a simple MATLAB® function by using GPU Coder™. A Mandelbrot set implementation by using standard MATLAB commands acts as the entry-point function. This example uses the codegen command to generate a MEX function that runs on the GPU. You can run the MEX function to check for run-time errors.

Open Live Script

Code Generation for Deep Learning Networks

Get started with CUDA code generation for image classification networks such as ResNet.

Open Live Script