High Performance Computing: Achieving Performance and Efficiency
In this course, developed in partnership with IEEE Future Directions, we try to give the context of current and near-future CPUs and GPUS and what developers are facing and will be facing in the near future when they go to program them. We discuss which programming languages dominate the field of HPC and which new ones are making their appearance. We also discuss parallel programming models since the architectures of today have very different kinds of parallelism that all can be and usually need to be exploited. Last, we set the stage for what programmers will have to deal with regarding workload and workflow management in many large-scale HPC environments by examining workloads, workflows, containers and why.
What you will learn:
- Discuss the hardware that makes everything happen
- Examine the software languages, models and tools for writing applications
- Consider Workload and Workflow Management & the necessary orchestration
This course is part of the following course program:
High Performance Computing: Technologies, Solutions to Exascale Systems, and Beyond
Courses included in this program:
Instructors
Barbara Chapman
Dr. Barbara Chapman obtained her Ph.D. in Computer Science from Queen’s University of Belfast and today is a Distinguished Technologist for the Cray Programming Environment at Hewlett Packard Enterprise (HPE). She has taught as a Professor of Computer Science at Stony Brook University for over 20 years and remains affiliated with the Department of Computer Science and the Institute for Advanced Computational Science. Her research group contributes to the open-source LLVM compiler infrastructure, especially to its OpenMP implementation, and has provided a reference implementation of the library-based OpenSHMEM parallel programming standard. Barbara has also served as Chair of the Department of Computer Science and Mathematics at the Department of Energy’s Brookhaven National Laboratory. Her research interests focus on parallel programming interfaces and related implementation technology and has authored more than 300+ refereed technical publications.
Larry Kaplan
Larry Kaplan earned his Bachelor’s degree in Computer Science modified with Electrical Engineering from Dartmouth College and a Master’s degree in Computer Architecture from NYU. Since that time, for the next 30 years, he has been working on HPC systems and played a significant role in the design of the XC, XE, XT, XMT, and MTA supercomputers, especially in the areas of runtime, operating and supervisory systems, and hardware/software interface. Past projects have included work in system and application resiliency, virtual and physical memory management, and network communication and management. Today Larry is a Senior Distinguished Technologist at HPE working as the Chief Software Architect for HPC focusing on the design and implementation of the software stack for the current and next generations of the HPE Cray Supercomputer product line. Special focus areas include federated workflows, advanced systems, and power management, open and modular software ecosystems, and advanced network software stacks from managers and drivers up through programming models.. Larry Kaplan holds over a dozen U.S. patents.
Publication Year: 2023
ISBN: 978-1-7281-7829-5