ICER

Led by Craig Grosch

project image

Institute for Cyber-Enabled Research (ICER)

The Institute for Cyber-Enabled Research (ICER) is the home to Michigan State University’s supercomputer, also known as the high-performance computing center (HPCC). Researchers and students across campus use the HPCC for performing a variety of computational tasks including large-scale simulations, artificial intelligence and machine learning, data analysis, and more!

The ICER VIP team will work on a new research project to help ICER quantify the environmental impact of its high-performance computing (HPC) workloads and explore ways in which that impact might be reduced. HPC often requires a significant amount of energy and cooling power, and as such, can negatively affect the environment via carbon emissions or excessive water usage. The goal of this project is to fully understand the sources contributing to carbon emissions in ICER’s HPCC and help make recommendations to reduce them.

Issues Involved or Addressed

The overarching goals of this research project are two-fold:

  • Quantify the emissions and efficiency of MSU’s supercomputer
    • Overall
    • On a per-workload basis
    • In comparison to other large-scale computing facilities
  • Make recommendations on how to reduce emissions
    • In hardware purchasing
    • In system tuning and scheduling
    • In user behavior

Methods and Technologies

As a researcher on this team, your work may cover a variety of tasks, including but not limited to:

  • Searching for and reading peer-reviewed resources to learn about sustainable computing.
  • Gathering data related to the HPCC, energy usage, and carbon efficiency. This may require generating your own data or finding sources that have the information you need.
  • Analyzing data using basic statistical formulas and potentially more advanced modeling techniques.
  • Producing written communications (e.g., blog posts or peer-reviewed publications) with clear and attractive visualizations.
  • Advertising and communicating your work through a variety of means including conferences.
  • Optionally helping develop new software or dashboards to improve user outcomes or visualize results.

Not every researcher will take part in every task, but all researchers will work together to accomplish the goals of the team.

While not every task will require all of the technologies below, the following are expected to be used somewhere in the project:

  • Git and GitLab for file version control and collaboration
  • Markdown files for documentation
  • High-performance computing environments including the Linux command line
  • Programming languages like Python or R (often via Jupyter or RStudio)

Please note that you are not expected to have prior experience with these technologies to join the team. We will teach these skills as necessary either through workshops, one-on-one discussions, or by providing external resources.

Areas of Interest

  • Computer science/engineering
  • Data science
  • Environmental science/engineering
  • Biosystems engineering
  • Sustainability
  • Artificial intelligence
  • Supply chain management
  • Education
  • Journalism
  • Anyone other majors interested in the project!

Preferred Interests and Preparation

The ICER team welcomes students from all academic backgrounds and levels of experience. This will be an interdisciplinary project with opportunities for contributions for students at many different stages in their academic career.

Specific skills and interests that are most relevant for this project include:

  • Interest in high-performance/research computing
  • Interest in gathering and/or analyzing data
  • Strong writing and communication skills
  • Strong willingness to collaborate

Other skills that may be used at some point in the project include:

  • Basic data analysis, visualization, and modeling skills
  • Basic programming skills (Python or R)
  • Some experience with the Linux command line

Do not worry if you don’t have experience with the skills listed above! We will be happy to teach them to anyone who is excited to learn.

Meeting Schedule & Location

  • Meeting Time: Wednesdays from 4:00-4:50PM
  • Location: 1442 BPS (or virtual in special circumstances)

Team Advisor

Dr. Craig Grosch
Dr. Craig Grosch
Institute for Cyber-Enabled Research (ICER), Michigan State University
grosscra@msu.edu

Dr. Grosch is a research computing facilitator at ICER and works to make research computing more accessible. His background is in applied mathematics and computer science, and his Ph.D. research is in developing fast and provably correct algorithms for solving high-dimensional problems. He is very excited to start a new research direction in sustainable computing!