Cross-Platform Performance Engineering with Arm Allinea Studio

John C. Linford; Principal Applications Engineer: Arm HPC Ecosystem
ALCF Developer Sessions: May 2019

This webinar will introduce Arm Allinea Studio, a cross-platform tool suite that supports C/C++, Fortran, and Python applications with hundreds of thousands of concurrent MPI processes and OpenMP threads executing on x86, POWER, NVIDIA GPUs, and Arm architectures. Cases studies demonstrating how MAP and DDT helped resolve issues at scale in simulations of quantum collisions and kinetic simulation of plasma will be presented. We will also demonstrate how Arm Allinea Studio can be used across multiple architectures at a variety of scales. The webinar will cover all components of Arm Allinea Studio, including Arm's compilers and scientific computing libraries, the DDT scalable debugger, and the MAP scalable profiler. Scientists and software engineers need intuitive, scalable, powerful performance engineering tools since software bugs are easily created, but hard to resolve, in the complexity of applications involving MPI, OpenMP, and GPUs. Most debuggers and profilers operate well on a small number of processes, but many bugs -- especially performance bugs -- only manifest at scale. This webinar will jumpstart new Arm Allinea Studio users to maximize the potential benefits of this toolkit.

About the Speaker

Dr. Linford has extensive experience creating, using, supporting, and deploying high-performance computing applications and technologies. His research interests include emerging computer architectures, compilers, code generation, performance analysis, and numerical simulation (particularly atmospheric chemistry). He has developed tools for extreme-scale software performance analysis, chemical kinetic simulation, rotorcraft engineering, and software environment management.