In 1965, the US and different nuclear powers dedicated to the Comprehensive Nuclear-Test-Ban Treaty, which prohibited nuclear exams. The National Nuclear Security Administration (NNSA), a successor to the Manhattan Undertaking, now exams nukes solely in simulation. To that finish, the NNSA yesterday unveiled the world’s quickest supercomputer to air in its mission to keep up a secure, safe, and dependable nuclear stockpile.
El Capitan was introduced yesterday on the SC Conference for supercomputing in Atlanta, Georgia, and it debuted at #1 within the newest Top500 list, a twice-yearly rating of the world’s highest performing supercomputers. El Capitan, housed at Lawrence Livermore National Laboratory in Livermore, Calif., can carry out over 2700 quadrillion operations per second at its peak. The earlier file holder, Frontier, might do exactly over 2000 quadrillion peak operations per second.
Alongside El Capitan, the NNSA introduced its unclassified cousin, Toulumne, which debuted at #10 on the Top500 listing and might carry out a peak of 288 quadrillion operations per second.
The NNSA—which oversees Lawrence Livermore in addition to Los Alamos National Laboratory and Sandia National Laboratories—plans to make use of El Capitan to “mannequin and predict nuclear weapon efficiency, getting old results, and security,” says Corey Hinderstein, appearing principal deputy administrator at NNSA. Hinderstein says the 3D modeling of a number of physics processes shall be considerably enhanced by the brand new supercomputer’s pace. The workforce additionally plans to make use of El Capitan to help in its inertial confinement fusion efforts, in addition to to coach artificial intelligence in help of each of these efforts.
Planning for El Capitan started in 2018, and building has been ongoing for the previous 4 years. The system is constructed by Hewlett Packard Enterprise, which has constructed the entire present high 3 supercomputers on the Top500 listing. El Capitan makes use of AMD’s MI300a chip, dubbed an accelerated processing unit, which mixes a CPU and GPU in a single package deal. In complete, the system boasts 44,544 MI300As, linked collectively by HPE’s Slingshot interconnects.
El Capitan makes use of 44,544 of AMD’s MI300A chips, which mix a CPU and GPU in a single package deal.Garry McLeod/Lawrence Livermore Nationwide Laboratory
Scientists are already at work porting their code over to the brand new machine, and they’re passionate about its promise. “We’re seeing vital pace ups in comparison with operating on previous chips versus this new factor,” says Luc Peterson, computational physicist at Lawrence Livermore Nationwide Laboratory. “We’re on the level the place our time to science is shrinking. We will do issues in just a few days that might have taken just a few months. So we’re fairly excited in regards to the purposes.”
But the urge for food for ever bigger supercomputers lives on. “We’re already engaged on the subsequent [high performance computing] acquisition,” says Thuc Hoang, director of the superior simulation and computing program at NNSA.
From Your Website Articles
Associated Articles Across the Internet
