Computer Architecture 101 for Scientists and Its Impact on Cluster Jobs

Dive into the fundamentals of supercomputer architecture and its impact on cluster job performance in this 48-minute talk by Dr. Georg Hager, Head of Training & Support at NHR@FAU. Explore essential concepts such as cores, sockets, caches, memory, networking, and data storage, and learn how understanding these elements can optimize your use of computational resources. Gain valuable insights into maximizing the efficiency of your CPU/GPU time allocation and demystify the complexities of hardware in high-performance computing. The presentation includes practical examples, such as GROMACS scaling analysis, and introduces tools like the HPC Portal and ClusterCockpit. Suitable for scientists and researchers looking to enhance their understanding of computer architecture and improve their cluster job performance.