This page provides you with documentation, training material and exercises relevant to POP users. Besides general training material regarding parallel programming, you can find information on our performance tools, explanatory material regarding our analysis methods and optimization techniques.
Performance Analysis
Documentation regarding POP Performance Audits
-
Original (POP1) Metrics
- Article explaining the POP Standard Metrics for Parallel Performance Analysis
- Presentation summarizing the POP Standard Metrics for Parallel Performance Analysis
-
New (POP2) Hybrid Metrics
- Introduction explaining the POP2 Standard Metrics for Performance Analysis of Hybrid Parallel Applications
- Cheat sheet for Additive Hybrid Metrics
- Cheat sheet for Multiplicative Hybrid Metrics
- In-depth explanation of the Additive Hybrid Metrics
- Webinar Identifying Performance Bottlenecks in Hybrid MPI + OpenMP Software (presentation slides)
-
How to create performance audits and assessments
- Article explaining how to create a POP performance audit yourself
- Introduction on how to obtain the POP Standard Metrics in Paraver
- Exercises and solutions how to start creating a POP performance audit yourself with Score-P/Cube/Scalasca or Extrae/Dimemas/Paraver
- Lightweight MPI Profiling with Intel MPI
- Lightweight MPI Profiling with mpiP
Documentation on how to understand POP performance tools displays shown in POP reports
Parallel Programming |
|
MPI: | MPI_Tutorial1; MPI_Tutorial2; MPI_Tutorial3; Exercises |
OpenMP: | IntroductionToOpenMP.pdf; OpenMPTaskingInDepth.pdf; OpenMPSummary.pdf; Exercises |
Performance Tools |
|
Introduction: | Parallel performance engineering |
Extrae: | Tutorial; UserManual |
Paraver: | Tutorial; UserManual |
Dimemas: | Tutorial |
Score-P: | Tutorial; UserGuide |
Cube: | Tutorial; UserGuide; DerivedMetrics |
Scalasca: | Tutorial; UserGuide; PerformanceProperties |
Exercise: | TutorialExerciseNPB; VI-HPS-LiveISO |
Darshan | ProfilingIOwithDarshan |
- POP Half-day Tutorial on Cube, Paraver and Darshan @ Supercomputing Frontiers Europe 2018
- VI-HPS Tools Guide
- State-of-the-art journal article "Scalable parallel performance measurement and analysis tools"
- State-of-the-art journal article "Tools for GPU Computing"