The Comedy Club of High-Performance Computing: Low-Rank Matrix Approximation Takes the Stage

Explore the world of high-performance computing in this engaging 55-minute NHR PerfLab seminar presented by Hatem Ltaief from King Abdullah University of Science and Technology (KAUST). Dive into the implementation of Tile Low-Rank Matrix-Vector Multiplication (TLR-MVM) on various hardware systems, a crucial computational kernel for seismic wave-equation-based processing and ground-based computational astronomy applications. Learn how TLR-MVM exploits data sparsity and utilizes efficient data layouts to maximize hardware performance. Discover the preliminary results showcasing TLR-MVM's superiority over dense implementations. Gain insights from Ltaief, a Principal Research Scientist specializing in parallel numerical algorithms, programming models, and performance optimizations for multicore architectures and hardware accelerators. Access the presentation slides and explore additional NHR PerfLab seminar events to further enhance your understanding of cutting-edge high-performance computing techniques.

The Comedy Club of High-Performance Computing: Low-Rank Matrix Approximation Takes the Stage!