Achieving Peak Performance of High-Performance Computing Applications by Optimizing Parallelism Compilation

Giorgis Georgakoudis | 21-ERD-018

Executive Summary

We aim to develop novel compiler methods to substantially increase the performance of high-performance computing applications running in parallel, addressing the problem of a severe lack of compiler optimization on parallel code. If successful, this technology will help enable the peak performance of applications on high-performance computing architectures, to speed stockpile stewardship simulations essential to national security.

Publications, Presentations, and Patents

Doerfert, Johannes, Marc Jasper, Joseph Huber, Khaled Abdelaal, Giorgis Georgakoudis, Thomas R. W. Scogland, and Konstantinos Parasyris. 2022. “Breaking the Vendor Lock — Performance Portable Programming Through OpenMP as Target Independent Runtime Layer.” PACT 2022. In Press.

Doerfert, Johannes, Atemn Patel, Joseph Huber, Shilei Tian, Jose M Monsalve Diaz, Barbara Chapman, and Giorgis Georgakoudis. 2022. “Co-Designing an OpenMP GPU Runtime and Optimizations for Near-Zero Overhead Execution.” 2022 IEEE International Parallel and Distributed Processing Symposium (IPDPS). May 2022, 504–14.

Georgakoudis, Giorgis, Thomas R. W. Scogland, Chunhua Liao, and Bronis R. de Supinski. 2022. “Extending OpenMP to Support Automated Function Specialization Across Translation Units.” OpenMP in a Modern World: From Multi-Device Support to Meta Programming, edited by Michael Klemm, Bronis R. de Supinski, Jannis Klinkenberg, and Brandon Neth, 159–73. Lecture Notes in Computer Science. Cham: Springer International Publishing, Sept. 2022.

Huber, Joseph, Melanie Cornelius, Giorgis Georgakoudis, Shilei Tian, Jose M. Monsalve Diaz, Kuter Dinel, Barbara Chapman, and Johannes Doerfert. 2022. “Efficient Execution of OpenMP on GPUs.” 2022 IEEE/ACM International Symposium on Code Generation and Optimization (CGO). April 2022, 41–52. IEEE.

Huber, Joseph, Weile Wei, Giorgis Georgakoudis, Johannes Doerfert, and Oscar Hernandez. 2021. “A Case Study of LLVM-Based Analysis for Optimizing SIMD Code Generation.” International Workshop on OpenMP, 142–55. Springer. Sept. 2022.

Jayatilaka, Tarindu, Hideto Ueno, Giorgis Georgakoudis, EunJung Park, and Johannes Doerfert. 2021. “Towards Compile-Time-Reducing Compiler Optimization Selection via Machine Learning.” 50th International Conference on Parallel Processing Workshop, 1–6. ICPP Workshops ’21. New York, NY, USA: Association for Computing Machinery.

Liao, Chunhua, Anjia Wang, Giorgis Georgakoudis, Bronis R. de Supinski, Yonghong Yan, David Beckingsale, and Todd Gamblin. 2021. “Extending OpenMP for Machine Learning-Driven Adaptation.” Accelerator Programming Using Directives: 8th International Workshop, WACCPD 2021, Virtual. November 14, 2021. Proceedings, 49–69. Berlin, Heidelberg: Springer-Verlag.

Mattson, Timothy G., Todd A. Anderson, and Giorgis Georgakoudis. 2021. “PyOMP: Multithreaded Parallel Programming in Python.” Computing in Science & Engineering 23 (6): 77–80, Nov. - Dec. 2021.

Parasyris, Konstantinos, Giorgis Georgakoudis, Johannes Doerfert, Ignacio Laguna, and Thomas R W Scogland. 2022. “Piper: Pipelining OpenMP Offloading Execution Through Compiler Optimization For Performance.” In Press.