XIAOYE (SHERRY) LI

Visible Spectrum is a series to spotlight talented and dedicated women employees across the Lab

March 15, 2022

Visible Spectrum is a series to spotlight talented and dedicated women employees across the Lab. Xiaoye (Sherry) Li is a senior scientist and the group leader of scalable solvers in the Applied Math and Computational Research Division. Her day-to-day responsibilities include research and development work on developing new mathematical algorithms, implementing on high-performance computing (HPC) machines, supervising junior scholars, working with the physical scientists and engineers who use the Lab’s math libraries.


In her spare time, Sherry enjoys staying active by running, biking, swimming, and playing volleyball.

What inspired you to work at Berkeley Lab?

The Lab has many amazing world-class researchers who have produced a tremendous amount of top-notch scientific breakthroughs and results that benefit society in so many different ways. The culture of open and team science at the Lab allows me to interact with numerous researchers from different groups and divisions to address new challenging scientific problems to be solved. That inspires me to always look for novel solution methodologies and research directions to tackle the biggest problems in science today.


What does your current scientific project or research entail?

I am leading several projects funded through the Exascale Computing Project (ECP) in developing sparse linear algebra algorithms and software for the Department of Energy's emerging HPC platforms, such as Perlmutter at NERSC. I also use statistical and machine learning tools for application code optimizations on these complex computer architectures.


What have you been most proud of in your work?

It is truly rewarding to see that the mathematical software I am developing is being widely used in multiple domains of science and engineering. For example, I have been the lead developer of the SuperLU library, which is one of the leading software libraries used for solving sparse linear systems of equations and has consistently been the most-downloaded software from the Berkeley Lab Computing Sciences organization. SuperLU has been used in designing new particle accelerators and modeling fusion energy reactors and in a variety of commercial applications, including (but not limited to): Walt Disney Feature Animation, in airplane designs, as well as the oil and semiconductor industries.


I have also been deeply involved in tackling fusion physics problems, which is focused on harnessing fusion and creating clean energy. Lab scientists are collaborating with those at Princeton Plasma Physics Laboratory to use computer modeling for a prototype fusion reactor, ITER, which is now being constructed in France. If successful, ITER will produce 10 times as much energy than it uses. In order to optimize the operation and energy output of ITER, scientists need to solve the complex equations that describe the magnetized plasma. Several of the leading codes that are being used for predicting and optimizing the performance of ITER have used the SuperLU library. This ability to make calculations of unprecedented accuracy and efficiency is critical to the success of the $10 billion ITER experiment.

Do you have tips you'd recommend for someone looking to enter and/or succeed in your field of work?

The field of high-performance computing with an emphasis on solving science and engineering problems is really diverse and interdisciplinary. Those who are most successful in this field usually have a very broad knowledge base in mathematics, computer science, statistics, and physical sciences. For graduate students who are looking to enter this field, I highly recommend taking foundational classes across different departments, and not only those restricted to their declared major. In addition, this field values both theoretical thinking as well as hands-on computer implementation skills.

How can our community engage more women, girls, and other underrepresented groups in STEM?

Empowerment needs to start from early on, even as early as kindergarten. If we can create and maintain an inclusive environment so that the girls and underrepresented communities can participate in activities side-by-side with others and collaborate with peers, this can help boost confidence and avoid limiting mindsets or prevent creating barriers based on gender and upbringing.