Cunxi Yu Receives NSF-DOE CS2 Award

A new project, “Formally Verified and Performance-Optimized Tensor Contraction Sequences in Quantum Many-Body Computations,” has been announced under the NSF–DOE Correctness for Scientific Computing Systems (CS2) program. The project focuses on improving both the correctness and performance of tensor contraction workflows that are central to quantum many-body computation.

Cunxi Yu, Assistant Professor in the Department of Electrical and Computer Engineering at the University of Maryland, brings his expertise in formal methods to this collaborative effort, which includes partners from North Carolina State University, Rutgers University, and DOE’s Pacific Northwest National Laboratory (PNNL).

“High-performance scientific computing is increasingly central to modern discovery, but performance alone is not enough,” said Yu. “We must also guarantee correctness and numerical reliability, especially in complex computational workflows such as tensor contractions for quantum many-body problems. Formal methods offer a rigorous foundation to both verify correctness and enable principled optimization, helping ensure that scientific results are not only efficient, but also trustworthy.”

Yu joined the UMD ECE Department in 2023 and has received the Best Paper Award at DAC 2023, the Best Paper Award at ASPLOS 2025, multiple best paper nominations, and the NSF CAREER Award in 2021.

Published April 21, 2026