Martin Karp

Martin Karp is a Research Fellow and postdoctoral researcher at KTH Royal Institute of Technology's Department of Engineering Mechanics, working under Dan Henningson. His research focuses on high-fidelity numerical simulations of turbulence and transition, with a specialization in high-performance computing (HPC) and supercomputing architectures. He holds a PhD in computer science from KTH and an MSc in Engineering Physics from Lund University, complemented by studies at ETH Zürich's computer science department.

His research interests explore computational limits in nonlinear chaotic systems and future computational advancements. He leads the development of the Neko framework, a scalable simulation tool for extreme-scale CFD with extensive accelerator support. Karp's work emphasizes GPU and FPGA acceleration, parallel computing, and optimizing algorithms for heterogeneous architectures.

Key contributions include large-scale turbulence simulations using GPUs, reducing communication in conjugate gradient methods, and evaluating FPGA-based flow solvers. His publications span journals like Concurrency and Computation and Scientific Reports, with conference presentations at IEEE Cluster, PASC, and HPCAsia. Karp's research bridges theoretical computational limits and practical HPC implementation, addressing challenges in precision, scalability, and hardware utilization.

His educational background combines engineering physics with computer science, enabling interdisciplinary approaches to fluid dynamics and high-performance simulation. Current projects aim to push the boundaries of computational fluid dynamics through novel algorithm design and leveraging emerging hardware capabilities.