Time-energy

The repository contains time-energy data and analysis for modern multicore systems, as presented in the research paper "The Time and Energy Efficiency of Modern Multicore Systems" (PARCO 2019). It also includes scripts and configurations for generating model data for homogeneous and heterogeneous multicore systems.

For homogeneous systems, the script `model_homogeneous.py` uses equations from `model.py` and takes parameters such as the number of cores, active power fraction (APF), idle system power, application name, and measured data. Configuration files for various homogeneous systems (AMD, ARM, Xeon, i7, Pi3) are provided.

For heterogeneous systems, the script `model_heterogeneous.py` uses equations from `model.py` and requires parameters like the total number of cores, number of little and big cores, APF for little and big cores, idle system power, application name, measured data for different core configurations and scheduling methods. Configuration files for heterogeneous systems (XU3, TX2) are provided.

Based on the research findings presented in the associated paper, the repository aims to facilitate the analysis and modeling of time-energy efficiency in modern multicore systems, both homogeneous and heterogeneous.

Topic

Data acquisition;Sample collections;Experimental design and studies

Detail

  • Operation: -

  • Software interface: Command-line user interface

  • Language: Shell,Python

  • License: Not stated

  • Cost: Free of charge

  • Version name: -

  • Credit: The Ministry of Education - Singapore through the Academic Research Fund Tier 1.

  • Input: -

  • Output: -

  • Contact: Dumitrel Loghin dumitrel@comp.nus.edu.sg

  • Collection: -

  • Maturity: -

Publications

  • Time-energy measured data on modern multicore systems running shared-memory applications.
  • Loghin D and Teo YM. Time-energy measured data on modern multicore systems running shared-memory applications. Time-energy measured data on modern multicore systems running shared-memory applications. 2019; 27:104670. doi: 10.1016/j.dib.2019.104670
  • https://doi.org/10.1016/J.DIB.2019.104670
  • PMID: 31709289
  • PMC: PMC6833352

Download and documentation


< Back to DB search