Register training material
2 materials found

Keywords: AMD 


AMD Profiling

The AMD profiling workshop covers the AMD suite of tools for development of HPC applications on AMD GPUs.

You will learn how to use the rocprof profiler and trace visualization tool that has long been available as part of the ROCm software suite.

You will also learn how to use the new...

Keywords: supercomputing, performance, GPUs, CPUs, AMD, HPC, ROCm

Resource type: activity

AMD Profiling https://dresa.org.au/materials/amd-profiling The AMD profiling workshop covers the AMD suite of tools for development of HPC applications on AMD GPUs. You will learn how to use the rocprof profiler and trace visualization tool that has long been available as part of the ROCm software suite. You will also learn how to use the new Omnitools - Omnitrace and Omniperf - that were introduced at the end of 2022. Omnitrace is a powerful tracing profiler for both CPU and GPU. It can collect data from a much wider range of sources and includes hardware counters and sampling approaches. Omniperf is a performance analysis tool that can help you pinpoint how your application is performing with a visual view of the memory hierarchy on the GPU as well as reporting the percentage of peak for many different measurements. training@pawsey.org.au supercomputing, performance, GPUs, CPUs, AMD, HPC, ROCm
Porting the multi-GPU SELF-Fluids code to HIPFort

In this presentation by Dr. Joseph Schoonover of Fluid Numerics LLC, Joe shares their experience with the porting process for SELF-Fluids from multi-GPU CUDA-Fortran to multi-GPU HIPFort.

The presentation covers the design principles and roadmap for SELF and the strategy to port from...

Keywords: AMD, GPUs, supercomputer, supercomputing

Resource type: presentation

Porting the multi-GPU SELF-Fluids code to HIPFort https://dresa.org.au/materials/porting-the-multi-gpu-self-fluids-code-to-hipfort In this presentation by Dr. Joseph Schoonover of Fluid Numerics LLC, Joe shares their experience with the porting process for SELF-Fluids from multi-GPU CUDA-Fortran to multi-GPU HIPFort. The presentation covers the design principles and roadmap for SELF and the strategy to port from Nvidia-only platforms to AMD & Nvidia GPUs. Also discussed are the hurdles encountered along the way and considerations for developing multi-GPU accelerated applications in Fortran. SELF is an object-oriented Fortran library that supports the implementation of Spectral Element Methods for solving partial differential equations. SELF-Fluids is an implementation of SELF that solves the compressible Navier Stokes equations on CPU only and GPU accelerated compute platforms using the Discontinuous Galerkin Spectral Element Method. The SELF API is designed based on the assumption that SEM developers and researchers need to be able to implement derivatives in 1-D and divergence, gradient, and curl in 2-D and 3-D on scalar, vector, and tensor functions using spectral collocation, continuous Galerkin, and discontinuous Galerkin spectral element methods. The presentation discussion is placed in context of the Exascale era, where we're faced with a zoo of available compute hardware. Because of this, SELF routines provide support for GPU acceleration through AMD’s HIP and support for multi-core, multi-node, and multi-GPU platforms with MPI. training@pawsey.org.au AMD, GPUs, supercomputer, supercomputing