The US Department of Energy (DoE) and Intel, in a recent press release statement on Monday, Mar. 18th 2019, announced plans to build the nation’s first exascale supercomputer at Argonne National Laboratory by 2021. The government agency claims it’ll be a major milestone that will bring a tectonic shift within the nation’s R&D, placing innovation, health research, and scientific discoveries on an accelerated path.
The new $500 million supercomputer, called Aurora, will be powered by the Intel Xeon processors and Cray hardware and will be the nation’s first high-performance computing system that touches exaflop speeds (a thousand petaflops, i.e. a billion billion calculations per second). Furthermore, the Aurora supercomputer will be at least seven times faster than the current Summit system, the fastest supercomputer in the world at the Energy Department’s Oak Ridge National Laboratory, which touches 143.5 petaflops speeds.
Speaking on a conference call, Rick Stevens, associate laboratory director for computing, environment, and life sciences at the Argonne National Laboratory, said: “This system will be an excellent platform for traditional high-performance computing applications, but it’s also being designed to be very fast for data analytics and an excellent platform for deep learning”.
Exascale speeds have been the supercomputing industry’s Holy Grail. The US, China, Japan, and several European countries all already into the race to achieve it.
Peter Rutten, research director of IDC’s core and edge infrastructure group, said: “The US, China, Japan, and Europe are in a race for achieving exascale computing, which is an artificial race. There are no losers. They will all get there, with perhaps a few months difference”.