George Biros

George Biros is a Professor of Mechanical Engineering and holds the W.A. 'Tex' Moncrief Jr. Endowed Chair in Simulation-Based Engineering Sciences at the University of Texas at Austin. He leads the Parallel Algorithms for Data Analysis and Simulation Group within the Institute for Computational Engineering and Sciences (ICES). His research focuses on computational mathematics, parallel algorithms for physics-based simulations, and large-scale data analysis, with applications in biomedical imaging, tumor growth modeling, and high-performance computing.

Biros earned his Ph.D. in Computational Science and Engineering from Carnegie Mellon University (CMU), followed by a postdoctoral fellowship at NYU's Courant Institute. He previously held faculty positions at the University of Pennsylvania and Georgia Tech. His work emphasizes scalable algorithms for exascale computing, including fast multipole methods and hierarchical matrix techniques.

Research interests span parallel algorithms, numerical methods for PDEs, inverse problems, and biophysical modeling. Notable contributions include tumor growth models with mass effect, image registration algorithms (e.g., CLAIRE), and GPU-accelerated solvers for fluid dynamics and plasma simulations. His group develops open-source libraries like PyKokkos for performance-portable Python kernels.

Recent publications highlight advancements in grain microstructure prediction via graph neural networks, single-scan MRI tumor calibration, and tau protein dynamics modeling in Alzheimer's disease. He has pioneered GPU-based diffeomorphic image registration for large biomedical datasets, achieving real-time performance on multi-GPU systems.

• Awards: Two-time ACM Gordon Bell Prize (2010, 2016), DOE Early Career Award
• Grants: Leads projects in exascale computing, numerical methods, and medical image analysis
• Labs: ICES Parallel Algorithms Group, ExaNIML (Exascale Numerically Inspired Machine Learning)