Lars Karlsson

Lars Karlsson is an Associate Professor at the Department of Computing Science, Umeå University, where he serves as Assistant Head of Department with responsibilities for undergraduate education. His research focuses on developing efficient algorithms for matrix and tensor computations within high-performance computing environments. He is affiliated with the Parallel and Scientific Computing research group at Umeå.

Karlsson's primary research areas include:

• Numerical Linear Algebra: Specializing in matrix factorizations and eigenvalue computations
• Parallel Algorithms: Designing scalable solutions for distributed and shared-memory systems
• Tensor Computations: Developing decomposition methods and completion algorithms
• Performance Optimization: Auto-tuning techniques for modern computing architectures

His publications (2010-2025) demonstrate consistent focus on parallel algorithms for numerical problems, particularly matrix reductions, eigenvalue computations, and tensor decompositions. Recent work emphasizes auto-tuning, robustness in numerical methods, and efficient scheduling for high-performance systems. The majority of publications involve collaborative research within European computing consortia like NLAFET.

Karlsson contributes to educational research, having explored mastery learning approaches in university settings. His administrative responsibilities include oversight of undergraduate programs at the Department of Computing Science.