Getting Started The ROCm Structure ROCm Courses Real World Applications

 

Getting Started
Learn about what ROCm is and how to install the software to get started and be set up for success.

Download the presentation

Watch the video

Read the documentation

Explore the ROCm Infinity Hub

 

The ROCm Structure

Applications HPC Applications ML Frameworks
Cluster Development Singularity SLURM Docker Kubernetes
Tools Debugger Profiler, Tracer System Valid. System Mgmt.
Portability Frameworks Kokkos RAJA GridTools ONNX
Math Libraries RNG, FFT Sparse BLAS, Eigen MIOpen
Scale-Out Comm. Libraries OpenMPI UCX
MPICH
RCCL
Programming Models
OpenMP HIP OpenCL
Processors CPU+GPU

ROCm Courses

Fundamentals of HIP

HIP is a high performance, CUDA-like programming model that is built on an open and portable framework. You will learn everything ranging from the basics of GPU programming to profiling GPU applications to porting your existing CUDA code, allowing you to run your applications on ROCm with ease.

Learn about programming with HIP

 

Multi-GPU Programming

Multiple GPU’s can be used to harness larger memory and attain greater speeds for an application. This module deepens your knowledge of ROCm™ and scaling across platforms through hands-on labs, examples, and trainings.

Learn about Multi-GPU Programming

Deep Learning with ROCm

Deep learning is part of a broader family of machine learning methods based on artificial neural networks with representation learning. This module includes hands on training for Deep Learning and equips you with the necessary knowledge on optimal usage of ROCm™ based systems.

Learn about Deep Learning

 

Real World Applications

Learn more about how ROCm is being used in real world research and development applications.

Watch the information sessions