Skip to content

This repository provides all the necessary resources to replicate and understand the simulations, experiments, and results presented in the paper: "Hybrid DRAM and Persistent Memory Architectures for Energy-Efficient and Scalable Big Data Management".

License

Notifications You must be signed in to change notification settings

habiiibo03/Hybrid-DRAM-and-Persistent-Memory-Architectures

Repository files navigation

VA-HMA Reproducibility Bundle

This repository bundle provides configuration files, scripts, and example datasets to reproduce the experiments in Volatility-Aware Hybrid Memory Architecture for Real-Time and Persistent Big Data Systems.

Contents

  • simulators/ — DRAMSim3 and NVMain config templates and driver stubs.
  • workloads/ — small synthetic workload generators (YCSB-like and Spark trace generator stub).
  • ml/ — feature extractor and Random Forest training script (scikit-learn).
  • analysis/ — result aggregation and plotting scripts.
  • supplementary/ — Supplementary tables of parameters and baseline configs.
  • run_all.sh — top-level script to run a minimal end-to-end example (requires external simulators).

Important This bundle provides ready-to-run scripts assuming the environment has:

  • DRAMSim3 installed and available on PATH (or you can provide the path in simulators/README_simulators.md).
  • NVMain installed and available on PATH.
  • Python 3.8+ with dependencies listed in requirements.txt (scikit-learn, numpy, pandas, matplotlib).
  • Apache Spark installed and configured if you want to run the Spark-based workloads.

See simulators/README_simulators.md for setup instructions and the run_all.sh script for an example workflow.

About

This repository provides all the necessary resources to replicate and understand the simulations, experiments, and results presented in the paper: "Hybrid DRAM and Persistent Memory Architectures for Energy-Efficient and Scalable Big Data Management".

Resources

License

Stars

Watchers

Forks

Packages

No packages published