GPU Coder Interface for Deep Learning

Use GPU Coder to generate optimized CUDA code for deep learning networks

Note: Starting in R2026a, to generate CUDA code directly from PyTorch and LiteRT Models, use the MATLAB Coder Support Package for PyTorch and LiteRT Models. To generate CUDA code from models using Deep Learning Toolbox, use this support package.

GPU Coder generates optimized CUDA code from MATLAB code and Simulink models for deep learning, embedded vision, and autonomous systems. You can deploy a variety of pretrained deep learning networks such as YOLOv2, ResNet-50, SegNet, MobileNet, and others from Deep Learning Toolbox to NVIDIA GPUs. You can generate optimized code for pre-processing and post-processing along with your trained deep learning networks to deploy complete applications.

• Supported deep learning networks

When used with GPU Coder, GPU Coder Interface for Deep Learning provides the ability for the generated code to call into cuDNN or TensorRT optimization libraries for NVIDIA GPUs.

When used in MATLAB with Deep Learning Toolbox and without GPU Coder, you can accelerate the execution of deep learning networks on NVIDIA GPUs.

This support package is functional for R2018b and beyond.

If you have download or installation problems, please contact Technical Support - https://www.mathworks.com/support/contact_us.html

Deep Learning on Jetson AGX Xavier using MATLAB, GPU Coder, and TensorRT