Course


Course Overview

AMD HIP Training Course Overview

Introduction

This course will prepare the students with the skills and knowledge required to use AMD ROCm platforms for high-performance computing applications. Students will learn how to write GPU programs using the HIP programming language, the default programming language for AMD GPUs on the ROCm platform. In addition, we highlight the skills needed to debug, profile, and improve the performance of HIP programs. This course will also cover how to convert existing CUDA programs to run on AMD GPUs. We introduce ROCm-provided libraries that can enable you to develop high performance applications and leverage multi-GPU scaling.

Course overview

  • Introduction to the HIP/ROCm
  • HIP Programming
  • AMD GPU Architecture
  • HIP Tools
  • Performance Tuning for HIP Programs
  • HIPify
  • HIP/ROCm Libraries
  • Multi-GPU Scaling

This course was originally delivered as a collaborative effort between AMD, Northeastern University, and The College of William and Mary, reproduced here with permission.