Skip to content

Sabya2/MPI-and-OpenMPI-based-PingPongLatency

Repository files navigation

Project Overview

This project implements a data transfer program using OpenMPI and IntelMPI, coded in C, for communication between compute nodes. The program is managed by the SLURM workload manager. Additionally, the project includes a Python pipeline designed for quantitative analysis and benchmarking of transfer data.

Features

  • C-Based Data Transfer: Implemented with both OpenMPI and IntelMPI for efficient communication between compute nodes.
  • SLURM Integration: Utilizes SLURM for managing and scheduling tasks across the compute cluster.
  • Python Analytics: A comprehensive pipeline for analyzing and benchmarking transfer data, providing valuable insights into performance metrics.

Combined Result

Code Layout

  • Analysis_python_transferData: Contains Python scripts dedicated to data transfer analytics and benchmarking.
  • PingPongLatency: Root directory containing the C-based data transfer code. intelMPI Code, openMPI Code,
    • intelMPI / openMPI: C-based data transfer code utilizing IntelMPI or OpenMPI packages.
      • intelCompiler / gnuCompiler: Used for compiling the C code with either Intel's compiler or GCC (GNU Compiler Collection).

About

C based code for data transfer between compute nodes

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages