Mansi Sakarvadia and Michael Tynes, two students who will enter the UChicago CS PhD program this fall, received the Department of Energy Computational Science Graduate Fellowship (DOE CSGF).
Sakarvadia and Tynes are two of 33 total fellows in the 2022 cohort of the program, which is meant to train future leaders in using high-performance computing for research across a range of fields. Each fellow receives a stipend, tuition support, and other benefits, including a three-month practicum opportunity conducting research at a DOE laboratory. Since the program was established in 1991, nearly 600 fellows have participated.
“[The] Office of Science is proud to support the training of a diverse and accomplished group of students to become leaders among the next generation of computational scientists,” said Barbara Helland, DOE Associate Director of Science for Advanced Scientific Computing Research. “As evidenced by the success of the current CSGF alumni, the new fellows’ research will advance efforts in a wide range of science and engineering topics that benefit Administration priorities and the American people.”
Both Sakarvadia and Tynes have previous experience working at DOE national laboratories and researching a wide variety of scientific topics. Sakarvadia has applied machine learning to studies in medicine, neuroscience, and marine biology, and worked on high-performance computing and data science at Argonne National Laboratory with Kyle Felker and Taylor Childers. Tynes did research in X-ray crystallography and psychology before his current position at Los Alamos National Laboratory, where he has spent two years exploring “autonomous discovery” for applications in chemistry and materials science.
“Autonomous discovery is basically when you have a machine learning system decide which science experiments or simulations are the most interesting or promising to do, and then it either tells a robot to do those experiments or launches those simulations on a high-performance computer,” Tynes said. “And then when it gets the results back, it takes those into account when deciding what to do next. You let it keep doing that until it optimizes some predefined target or finds something new and unexpected.”
Working in this subject area led Tynes to discover Globus Labs, the research group of UChicago CS Professor Ian Foster and Research Associate Professor Kyle Chard. Similarly, Sakarvadia came across Foster and Chard’s work when looking for computer science labs with a broad focus on both questions within CS and applications in other disciplines.
“I knew that I really enjoyed research in general, and I had tried a lot of different things. I didn’t know at the time exactly what topic I wanted to do in my PhD, I just sort of broadly knew I enjoyed computer science,” Sakarvadia said. “So I was trying to find a lab that had an intersection of lots of different research topics and still encompassed some foundational corners of computer science. I got the sense that Globus Labs was really diverse in the types of projects they worked on, so I decided to apply to it. It seemed like a good fit for some of the ideas I had.”
In addition to the support for their graduate education, both students said they were excited about the fellowship opportunities to grow within the Department of Energy.
“It provides a really good networking opportunity to learn more about what’s going on in the DOE and to meet people who are interested in what you’re working on, or who you may be interested in working with,” Tynes said. “They really want you to broaden your experience and give you the opportunity to integrate yourself into the DOE.”