CS 74 — An Introduction to Building High Performance Computing Systems
Course Format: Live Online (About Formats)
Duration: 8 weeks
Date(s): Oct 4—Nov 29
Time: 7:00—8:50 pm (PT)
Refund Deadline: Oct 6
Instructor(s): Steve Jones
Class Recording Available: Yes
High performance computing (HPC) is a field at the forefront of a range of high tech applications such as computational fluid dynamics, image processing, and financial risk management. With the demands of machine learning outstripping conventional computing, HPC is also at the forefront of artificial intelligence. This course will discuss how HPC clusters are used in large-scale problems in academia and industry alike. Students will learn about HPC clusters from the ground up and gain a solid foundation in parallel computer architectures, cluster operating systems, and resource management. They will build their own systems via remote installation of physical hardware, configuration and optimization of a high-speed network, and integration of other technologies used throughout the HPC world. Discussion and collaboration will be key components of the course. Students will come away with a solid skill set in a field of computing that has broad implications for science and technology.
No knowledge of advanced computing systems, programming languages, or Linux is required.
Steve Jones has managed large-scale HPC clusters at Stanford since 2002 and leads the Stanford HPCC team of undergraduate students, operating three production clusters providing over 350 million core hours each year. He teaches Stanford's "Introduction to High Performance Computing," presents an HPC AI Seminar Series, is the co-chair of the Stanford HPC-AI Advisory Council Conference, and operates an HPC cluster-building academy with the HPC-AI Advisory Council.
Director, High Performance Computing Center, and Research Scientist, Stanford; Founder, HPC Clusters
Textbooks for this course:
There are no required textbooks; however, some fee-based online readings may be assigned.