target audience: TECH SUPPLIER  Publication date: Jul 2022 - Document type: Market Perspective - Doc  Document number: # US49300922

The Next "Frontier" in Supercomputing — U.S. Department of Energy's Oak Ridge National Laboratory's Exascale System Takes Top Spot on the 2022 "TOP500" List

By: 

  • Josephine Palencia
  • Peter Rutten Loading
  • Ashish Nadkarni Loading

Content



Get More

When you purchase this document, the purchase price can be applied to the cost of an annual subscription, giving you access to more research for your investment.



Related Links

Abstract


This IDC Market Perspective provides details about the exascale system of the Oak Ridge National Laboratory, U.S. Department of Energy, that takes top spot on the 2022 "TOP500" list. Impressive achievement with the successful collaboration of U.S. DOE's ORNL, HPE, and AMD pays off after three years of painstaking work, ushering in the first-ever, fastest, exascale supercomputer in the world. With the powerful leap in computational ability, breakthroughs in critical fields such as cancer research, accelerated drug discovery, renewable energy, and new materials research will pave the way to solve the world's most challenging problems, marking a new era in scientific computing. With this impressive pioneering effort, the rest of the computing community of vendors and large businesses will follow suit, including but not limited to NVIDIA, Graphcore, Cerebras, and Groq. Performance, cost, and affordability (full TCO and ROI analysis) are the dominant factors that determine which infrastructure national laboratories and other large businesses will choose to adopt. Other decision points include power usage efficiency and infrastructure footprint. With progress made toward more enhanced computing capability, a new paradigm also needs to be established to manage the scientific data, which has been ceaselessly exploding at a rate faster than the compute capability.

"Supercomputing centers and large businesses that enter into the realm of exascale computing will produce copious amounts of data at exponential rates reaching yottabyte (1024) and xenottabyte (1028) scales and even beyond," says Josephine Palencia, research director, High Performance Computing at IDC. "For data management, clear and balanced policies for data security and compliance are necessary to increase, facilitate, and allow open share access to sensitive data, which is problematic in the world of global hacking. These present opportunities for visionary and technological breakthrough innovations in HPC storage and data management in the same light."



Coverage


Do you have questions about this document
or available subscriptions?