How To Kill A Supercomputer: Dirty Power, Cosmic Rays, and Bad Solder

Will future exascale supercomputers be able to withstand the steady onslaught of routine faults?

11 min read
How To Kill A Supercomputer: Dirty Power, Cosmic Rays, and Bad Solder
Illustration: Shaw Nielsen

As a child, were you ever afraid that a monster lurking in your bedroom would leap out of the dark and get you? My job at Oak Ridge National Laboratory is to worry about a similar monster, hiding in the steel cabinets of the supercomputers and threatening to crash the largest computing machines on the planet. 

The monster is something supercomputer specialists call resilience—or rather the lack of resilience. It has bitten several supercomputers in the past. A high-profile example affected what was the second fastest supercomputer in the world in 2002, a machine called ASCI Q at Los Alamos National Laboratory. When it was first installed at the New Mexico lab, this computer couldn’t run more than an hour or so without crashing.

Keep reading...Show less

This article is for IEEE members only. Join IEEE to access our full archive.

Join the world’s largest professional organization devoted to engineering and applied sciences and get access to all of Spectrum’s articles, podcasts, and special reports. Learn more →

If you're already an IEEE member, please sign in to continue reading.

Membership includes:

  • Get unlimited access to IEEE Spectrum content
  • Follow your favorite topics to create a personalized feed of IEEE Spectrum content
  • Save Spectrum articles to read later
  • Network with other technology professionals
  • Establish a professional profile
  • Create a group to share and collaborate on projects
  • Discover IEEE events and activities
  • Join and participate in discussions

Q&A: Marc Raibert on the Boston Dynamics AI Institute

The founder of Boston Dynamics talks with us about the new $400 million research institute

12 min read
Marc Raibert, an older white man with a bald head and a short white beard and glasses, gestures as he speaks on a stage. He is wearing formal pants and a flower-print short sleeve shirt.

Marc Raibert, founder and chairman of Boston Dynamics, speaks at a Hyundai Motor Group news conference during CES 2022 in Las Vegas, Nevada.

Steve Marcus/Reuters/Alamy

Last week, Hyundai Motor Group and Boston Dynamics announced an initial investment of over $400 million to launch the new Boston Dynamics AI Institute. The Institute was conceptualized by (and will be led by) Marc Raibert, the founder of Boston Dynamics, with the goal of “solving the most important and difficult challenges facing the creation of advanced robots.” That sounds hugely promising, but of course we had questions—namely, what are those challenges, how is this new institute going to solve them, and what are these to-be-created advanced robots actually going to do? And fortunately, IEEE Spectrum was able to speak with Marc Raibert himself to get a better understanding of what the Institute will be all about.

Keep Reading ↓Show less

Emmy Award Winner’s Algorithms Bring High-Quality Video to Your TV

He is working on making high-res images for the metaverse

5 min read
portrait of Alan Bovik
Alan Bovik

Alan Conrad Bovik’s passion for science fiction inspired him to pursue a career in engineering. His favorite sci-fi authors when he was young were Arthur C. Clarke, who penned 2001: A Space Odyssey, and Isaac Asimov, author of the Foundation series. Bovik says they wrote from a “very scientific point of view”—which made him want to help develop aerospace technology that would send humans “to other worlds.”

But he decided to study nuclear engineering in school—which then seemed like the future of energy. He discovered, however, that he didn't like the subject because it “required too much chemistry and memorization,” he says with a laugh. When he took a course in computer programming, he fell in love with it and ended up changing his major to computer engineering.

Keep Reading ↓Show less

Harnessing the Power of Innovation Intelligence

Through case studies and data visualizations, this webinar will show you how to leverage IP and scientific data analytics to identify emerging business opportunities

1 min read
Clarivate
Clarivate

Business and R&D leaders have to make consequential strategic decisions every day in a global marketplace that continues to get more interconnected and complex. Luckily, the job can be more manageable and efficient by leveraging IP and scientific data analytics. Register for this free webinar now!

Join us for the webinar, Harnessing the power of innovation intelligence, to hear Clarivate experts discuss how analyzing IP data, together with scientific content and industry-specific data, can provide organization-wide situational awareness and reveal valuable business insights.

Keep Reading ↓Show less