Jul 07 2023
Hardware

Exascale Computing Takes Research to the Next Level

Exascale computing opens the door to higher ed research opportunities.

Exascale computing has arrived, offering a new way to help solve some of the world’s most complex problems. The lightning-fast performance of an exascale computer (one quintillion operations per second; “exa” refers to the number’s 18 zeros) has potential in research spaces that require incredibly large amounts of computer processing power. The country’s first exascale computers include Frontier at the Oak Ridge National Laboratory, Aurora at Argonne National Laboratory and El Capitan at Lawrence Livermore National Laboratory.

Before exascale, the fastest supercomputers in the world could handle problems at the petascale, or one quadrillion operations each second. Though petascale systems are still quite powerful, exascale marks a transformative change, allowing the study of systems that otherwise would be impractical or impossible to investigate in the real world due to their complexity, size, fleeting nature or potential risks.

Click the banner below to learn more about the technology behind today's HyFlex learning spaces.

Exascale Computing Accelerates Research Initiatives

Lori Diachin, the newly appointed director of the U.S. Department of Energy’s Exascale Computing Project (ECP), says that exascale computers will “more realistically simulate the processes involved in scientific discovery and national security,” including:

  • Precision medicine
  • Regional climate change
  • Additive manufacturing
  • The conversion of plants to biofuels
  • The relationship between energy and water use
  • The unseen physics in materials discovery and design
  • Fundamental forces of the universe

WATCH: How network upgrades are accelerating research at one university.

Exascale computing already has produced significant advances across distinct areas:

  • Modeling of wind turbines interacting in realistic settings, which can maximize energy output and minimize wear and tear on turbines over time
  • Using machine learning approaches to understand mechanisms behind RAS protein-based cancers (which account for 30 percent of all cancers) at a deep molecular level
  • Simulating a large-magnitude earthquake in the San Francisco Bay Area at a fine enough resolution that ground motion can be coupled to engineering codes that model buildings to assess the risk of earthquake damage

Exascale Computing’s Impact on University Research

High-performance computing is a critical component of university research programs. University researchers now have access to Energy Department exascale computers at the Oak Ridge, Argonne and Lawrence Livermore laboratories through peer-review allocation programs such as the Innovative and Novel Computational Impact on Theory and Experiment program and the ASCR Leadership Computing Challenge.

Universities also have access to the ECP’s Extreme-Scale Scientific Software Stack (E4S) project, which offers HPC developers a number of reusable HPC libraries and tools, including 100 HPC, artificial intelligence and high-performance data analytics-supported packages and commercial tools.

Diachin says that the ECP has been working closely with National Science Foundation centers. They are exploring implementation of E4S on the leadership class computers that will be deployed in the next few years at the NSF-funded Texas Advanced Computing Center at the University of Texas.

Exascale Computing Breakthroughs Could Impact Multiple Industries

By 2024, a total of three systems will be online at the Energy Department, according to Diachin. Over the next five to 10 years, Diachin predicts, the number of application areas that can leverage exascale computing will continue to expand to include an increasing number of users from university and industry centers.

DIG DEEPER: How HPE is advancing high-performance computing to meet rising demand.

Such an exponential increase in memory, storage and computing power will drive breakthroughs across all industries, including:

  • Energy production
  • Storage and transmission
  • Materials science
  • Additive manufacturing
  • Chemical design

“The hardware technologies and advances developed as part of the exascale computing initiative, such as extremely power-efficient computing node design and packaging, will continue to find their way to microelectronics available to the broader market, directly impacting large segments of the consumer market,” she says.

UP NEXT: How is higher education preparing for quantum computing?

Courtesy of Argonne National Laboratory
Close

Become an Insider

Unlock white papers, personalized recommendations and other premium content for an in-depth look at evolving IT