Q&A with Kristin Munch: Advancing Computing Capabilities To Efficiently Simulate, Analyze Complex Energy Systems
Feb. 26, 2021
Advanced computing is essential to research at the National Renewable Energy Laboratory (NREL). As laboratory program manager for Advanced Computing, Kristin Munch helps develop the vision and strategy for NREL’s computational infrastructure and ecosystem. She collaborates with researchers, leadership, and technology offices and programs in the U.S. Department of Energy Office of Energy Efficiency and Renewable Energy (EERE) to maximize the benefits of high-performance computing through improved robustness, agility, usability, and productivity.
She joined NREL in 2009 and served as the manager of the Computational Science Center’s Data, Analysis, and Visualization group for more than a decade before accepting her new role as laboratory program manager in October 2020.
What inspired you to pursue degrees in math, computer science, and computer engineering?
I majored in mathematics as an undergraduate student and then started working in the telecommunications industry. While working, I had the opportunity to pursue a master’s degree in computer engineering. When I finished my master’s, I decided to pursue a Ph.D.—I received a National Science Foundation grant, called the Integrative Graduate Education and Research Traineeship, at the University of Minnesota. That program emphasized interdisciplinary thesis work, which took me into the field of computational neuroscience. That work refocused my efforts in machine learning and data-intensive techniques of computer science. After finishing my Ph.D., I worked in a few industries, including pharmaceuticals, before joining NREL about 11 years ago. I wanted to do interdisciplinary work and wanted to be out West. (I’m originally from Maine.) The amount of research happening across so many domains at NREL makes it the perfect place for interdisciplinary work.
Let’s talk about NREL’s next supercomputer that will be hosted in the Energy Systems Integration Facility, or ESIF, which is one of your current projects.
We’re currently in the procurement phase for the supercomputer that will be replacing Eagle. This takes a few years and involves gathering computing requirements from users and EERE offices. We’re receiving proposals from multiple vendors and evaluating them with a broad technical team of computational researchers from across NREL, as well as computing technical advisors from the Materials, Chemical, and Computational Science (MCCS) directorate. We’ll have a contract in place this summer, and the next computer will be coming online in fall 2022.
We expect the new supercomputer will be a heterogeneous system comprised of both CPU’s and GPU’s. It will have more advanced processors, storage, and networking than Eagle within a similar footprint in the data center. We are also expecting the new computer to have more GPUs, or graphics processing units, than Eagle. With more GPUs, we can enhance our high-performance computing applications in material science, chemistry, fluid dynamics—and especially artificial intelligence and machine learning problems, which are becoming bigger parts of the EERE computing portfolio.
Another new capability coming down the pipe is a facelift to the Insight Center, or NREL’s visualization center. Can you tell me a little about that?
Yes, over the next two years, we’ll be updating the two-dimensional visualization room to have a new wall-size power display. That kind of display is incredibly useful for very large grid control and renewable scenario visualizations over large geographic regions—it’s an exciting new capability. In addition, we’re incorporating touch-screen displays that will look like standing monitors. Researchers will be able to touch the screen, move their data around, zoom in and out, or swipe to move their data back and forth between the display wall and the touch screens. This will make it easier to investigate huge amounts of data. The following year, we’ll upgrade the three-dimensional immersive room to have much higher resolution, improving our researchers’ experience walking through complex data structures like a multiscale energy system.
What do you look forward to when we can safely return to the NREL campus?
Most computer science work can be done remotely, so I feel lucky that my work hasn’t been dramatically impacted by the pandemic, but I definitely miss being on campus and seeing everyone. I’ll look forward to going to the Insight Center and different labs again and watching researchers build out capabilities.
What do you like to do outside of work?
I like running, hiking, and biking. But l prefer biking with a destination—like riding to work—not going on a country road ride for five hours. Every activity I do is to keep me in shape for running. I’ve been running for 35 years (since the sixth grade). I did cross-country and track in high school and cross-country in college. I’ve really enjoyed Colorado’s trail races that happen over the summer—they’re really fun, but they can be extremely challenging!