Using MPIProf for Performance Analysis
MPIProf is a lightweight, profile-based application performance analysis tool that works with many MPI implementations, including HPE MPT. The tool gathers statistics in a counting mode via PMPI, the...
View ArticlePorting with Intel-MPI
You can use the Intel MPI library to build and run your MPI application. The Intel MPI modules available on NAS resources are: mpi-intel/2018.3.222 mpi-intel/2019.5.281 mpi-intel/2020.0.166 For more...
View ArticleUsing Gprof for Performance Analysis
Gprof is a compiler-assisted performance profiler for C, Fortran, and Pascal applications running on Unix systems. You can use Gprof to help identify hotspots in your application where code...
View ArticleGetting a Quick Performance Overview with Intel APS
The Intel Application Performance Snapshot (APS) tool provides a quick overview of your application's performance on processor and memory usage, message passing interface (MPI), and I/O, as well as...
View ArticleUsing the IOT Toolkit for I/O and MPI Performance Analysis
IOT is a licensed toolkit developed by I/O Doctors, LLC, for I/O and MPI instrumentation and optimization of high-performance computing programs. It allows flexible and user-controllable analysis at...
View ArticleUsing Intel Advisor for Better Threading and Vectorization
You can use Intel Advisor to identify issues and help you with threading and vectorization optimization on the specific Intel processors that your Fortran, C, or C++ applications run on. The software,...
View ArticleRunning a Roofline Analysis with Intel Advisor
A roofline analysis helps you determine whether your application has achieved the best possible performance, limited by the machine capabilities. If not, you can explore the possibility of algorithm...
View ArticleUsing MPInside for Performance Analysis and Diagnosis
MPInside, an MPI application performance analysis and diagnostic tool from HPE, is a lightweight program that can be used with thousands of ranks. The tool is profile-based via PMPI, the MPI standard...
View ArticleOverview of Intel VTune Analysis
The Intel VTune Profiler is an analysis and tuning tool that provides predefined analysis configurations to address performance questions. This article provides basic information on several Intel VTune...
View ArticleFinding Hotspots in Your Code with the Intel VTune Command-Line Interface
The Intel VTune Profiler (renamed from Amplifier starting with 2020.0 version) is an analysis and tuning tool that provides predefined analysis configurations to address various performance questions....
View ArticleUsing Darshan for I/O Characterization
Darshan is an open-source, lightweight high-performance computing I/O characterization toolkit developed at Argonne National Lab. Darshan instruments applications and captures calls to the POSIX,...
View ArticleUsing the NAS Recommended MPT Library
UPDATE IN PROGRESS: Due to NASA security requirements, nearly all HECC systems will transition from the TOSS 3 operating system to TOSS 4 over the next month. Some versions of HPE MPT are not...
View ArticleProgramming for NVIDIA GPUs
UPDATE IN PROGRESS: The new Cabeus GPU system is currently comprised of 128 NVIDIA GPU A100 nodes, with plans to integrate the NVIDIA V100 nodes from Pleiades in early 2024. To develop code for use on...
View ArticleUsing the mbind Tool for Pinning
The mbind utility is a "one-stop" tool for binding processes and threads to CPUs. It can also be used to track memory usage. The utility, developed at NAS, works for for MPI, OpenMP, and hybrid...
View Article
