Petascale computing

Last updated

Petascale computing refers to high-power computing systems capable of performing at least 1 quadrillion (1015) arithmetic, or floating-point operations per second (FLOPS) [1] . (Note: FLOPS is a standard metric for supercomputer performance.) These systems are often called petaflops systems and represent a significant leap from traditional supercomputers in terms of raw performance, enabling them to handle vast datasets and complex computations.

Contents

Petascale computing typically finds application in large-scale applications like climate and weather modeling, astrophysics simulations, advanced materials design, and realtime medical imaging. [1] .

Definition

Floating point operations per second (FLOPS) is one measure of computer performance (including supercomputers). FLOPS can be recorded in different measures of precision, however the standard measure (used by the TOP500 supercomputer list) uses 64 bit (double-precision floating-point format) operations per second using the High Performance LINPACK (HPLinpack) benchmark. [2] [3]

The metric typically refers to single computing systems, although it can be used to measure distributed computing systems for comparison. It can be noted that there are alternative precision measures using the LINPACK benchmarks which are not part of the standard metric/definition. [3] It has been recognized that HPLinpack may not be a good general measure of supercomputer utility in real world application, however it is the common standard for performance measurement. [4] [5]

History

The petaFLOPS barrier was first broken by the RIKEN MDGRAPE-3 supercomputer in 2006, [6] [7] and then on 16 September 2007 by the distributed computing Folding@home project. [8] IBM's single petascale system, the Roadrunner, entered operation in 2008. [9] The Roadrunner, built by IBM, had a sustained performance of 1.026 petaFLOPS. The Jaguar became the next computer to break the petaFLOPS milestone, later in 2008, and reached a performance of 1.759 petaFLOPS after a 2009 update. [10]

In 2020, Fugaku became the fastest supercomputer in the world, reaching 415 petaFLOPS in June 2020. Fugaku later achieved an Rmax of 442 petaFLOPS in November of the same year.

In 2022, exascale computing (1018 FLOPS of computational power) overtook petascale computing in terms of power with the development of Frontier, surpassing Fugaku with an Rmax of 1.102 exaFLOPS in June 2022. [11]

As of January 2026, El Capitan is the world's fastest exascale computer. Built by Hewlett Packard Enterprise (HPE) for Lawrence Livermore National Laboratory (LLNL), Livermore, California, it runs at a groundbreaking speed of 1.809 exaflops. [12]

Artificial intelligence

Modern artificial intelligence (AI) systems require large amounts of computational power to train model parameters. OpenAI employed 25,000 Nvidia A100 GPUs to train GPT-4, using a total of 133 septillion floating-point operations. [13]

See also

References

  1. 1 2 "Research". researchgate.net. Retrieved January 19, 2026.{{cite web}}: CS1 maint: url-status (link)
  2. ↑ "FREQUENTLY ASKED QUESTIONS". www.top500.org. Retrieved 23 June 2020.
  3. 1 2 Kogge, Peter, ed. (1 May 2008). ExaScale Computing Study: Technology Challenges in Achieving Exascale Systems (PDF). United States Government. Retrieved 28 September 2008.
  4. ↑ Bourzac, Katherine (November 2017). "Supercomputing poised for a massive speed boost". Nature. 551 (7682): 554–556. doi: 10.1038/d41586-017-07523-y . Retrieved 3 June 2022.
  5. ↑ Reed, Daniel; Dongarra, Jack. "Exascale Computing and Big Data: The Next Frontier" (PDF). Retrieved 3 June 2022.
  6. ↑ Taiji, M.; et al. (2004), "MDGRAPE-3: A petaflops special-purpose computer system", Parallel Computing, Advances in Parallel Computing, vol. 13, Elsevier, pp. 669–676, doi:10.1016/s0927-5452(04)80083-2, ISBN   978-0-444-51689-3
  7. ↑ "Completion of a one-petaflops computer system for simulation of molecular dynamics" (PDF). Riken. June 19, 2006. Archived from the original (PDF) on 2014-09-08.
  8. ↑ Michael Gross (2012). "Folding research recruits unconventional help". Current Biology. 22 (2): R35–R38. doi: 10.1016/j.cub.2012.01.008 . PMID   22389910.
  9. ↑ National Research Council (U.S.) (2008). The potential impact of high-end capability computing on four illustrative fields of science and engineering. The National Academies. p. 11. ISBN   978-0-309-12485-0.
  10. ↑ National Center for Computational Sciences (NCCS) (2010). "World's Most Powerful Supercomputer for Science!". NCCS. Archived from the original on 2009-11-27. Retrieved 2010-06-26.
  11. ↑ "June 2022 | TOP500". www.top500.org. Retrieved 2024-11-21.
  12. ↑ "What is Exascale Computing? | Glossary | HPE India". hpe.com. Retrieved January 19,2026.{{cite web}}: Check date values in: |access-date= (help)CS1 maint: url-status (link)
  13. ↑ Minde, Tor Björn (2023-10-08). "Generative AI does not run on thin air". RISE. Retrieved 2024-03-29.
This page is based on this Wikipedia article
Text is available under the CC BY-SA 4.0 license; additional terms may apply.
Images, videos and audio are available under their respective licenses.