Evaluate Application Performance using TAU and E4S (April 4-5)
9:00am - 3:00pm
This workshop is offered hybrid: both in-person and virtual/online. It is a 2-day event, April 4-5, from 9:30am – 3:30pm AWST. There is a limited number of seats, so please register ONLY if you plan to attend both days.
This workshop is for computational scientists who want to evaluate the performance of their parallel, scientific applications.
In this workshop, you will learn about the Extreme-scale Scientific Software Stack and the TAU Performance System® and its interfaces to other tools and libraries.
During the hands-on portion of the workshop, the instructor will attempt to collect and analyse performance data for additional user codes. Attendees are welcome to contact the instructor ahead of time to begin collecting data to discuss at the workshop. (Contact firstname.lastname@example.org, providing more information.)
More Detail on the Workshop
To meet the needs of computational scientists to evaluate the performance of their parallel, scientific applications, we will focus on the use of E4S containers and TAU performance data collection, analysis, and performance optimization.
After demonstrating how performance data (both profile and trace data) can be collected using TAU’s (Tuning and Analysis Utilities) automated instrumentation, attention will then turn to how to instrument key codes and to analyse the performance data collected to explain where the time is spent.
The workshop will include sample codes that illustrate the different instrumentation and measurement choices. Topics will cover generating performance profiles and traces with memory utilization and headroom, I/O, and interfaces to ROCm, including ROCProfiler and ROCTracer with support for collecting hardware performance data.
The workshop will cover instrumentation of OpenMP programs using OpenMP Tools Interface (OMPT), including support for target offload and measurement of a program’s memory footprint. We will demonstrate scalable tracing using OTF2 and visualization using the Vampir trace analysis tool. Performance data analysis using ParaProf and PerfExplorer will be demonstrated using the performance data management framework (TAUdb) that includes TAU’s performance database.
Instructor: Sameer Shende
Prof. Sameer Shende has helped develop the TAU Performance System, the Program Database Toolkit (PDT), the Extreme-scale Scientific Software Stack (E4S) and the HPCLinux distro. His research interests include tools and techniques for performance instrumentation, measurement, analysis, runtime systems, HPC container runtimes, and compiler optimizations. He serves as a Research Professor and the Director of the Performance Research Laboratory at the University of Oregon, and as the President and Director of ParaTools, Inc. (USA) and ParaTools, SAS (France). He received his B.Tech in Electrical Engineering from IIT Bombay in 1991 and his M.S. and PhD in Computer and Information Science from the University of Oregon in 1996 and 2001 respectively.