This site may earn affiliate commissions from the links on this page. Terms of utilise.

IBM Summit

More supercomputer news this week: The U.s. is responding to China's new Sunway TaihuLight organisation that was announced Monday, and fast. First, the Department of Energy'south (DOE) Oak Ridge National Laboratory is expected to have delivery of a new IBM arrangement, named Summit, in early on 2022 that will now be capable of 200 superlative petaflops, Computerworld reports. That would get in nigh twice as fast as TaihuLight if the claim proves true. (We had originally reported in 2022 that both Peak and Sierra would reach roughly 150 petaflops.)

TaihuLight (pictured beneath) at present sits at number one on the twice-yearly TOP500 list of the fastest supercomputers in the world, with a Linpack benchmark score of 93 petaflops and a claimed peak of 124.5 petaflops. The latest TOP500 announcement Monday caused a bit of a stir. Non only is TaihuLight roughly three times faster than Mainland china's Tianhe-2, the prior champion, just it besides uses no United states-sourced parts at all for the start time, as information technology's powered past Sunway 260-cadre SW26010 processors that are roughly on par with Intel Xeon Phi, besides every bit custom proprietary interconnect.

sunway-taihulight

In plough, Meridian will employ IBM Power9 and Nvidia Volta GPUs. Summit will evangelize over five times the computational functioning of Titan's 18,688 nodes using merely well-nigh three,400 nodes. Each node will have "over half a terabyte" of then-called coherent memory (HBM + DDR4), plus 800GB of non-volatile RAM that serves as a flare-up buffer or extended memory.

Summit IBM

Titan (pictured beneath), meanwhile, is currently #iii on the TOP500 listing, and resides at the DOE Oak Ridge National Laboratory, where the new IBM Summit organisation will be located. In a argument via Computerworld, the DOE pointed out that since 1993, US supercomputing capabilities "accept grown exponentially by a factor of 300,000," and that "high-functioning calculating remains an integral priority for the DOE." (For more on Oak Ridge's earth, free energy, climate, and geographic science research using these systems, visit its dedicated supercomputing page.

Titan supercomputer

The Linpack criterion has emerged every bit a singular yardstick for measuring the performance of complex supercomputers. It doesn't record overall performance in all situations; rather, it measures the performance of a organization when solving a "dense organisation of linear equations" that gives a skillful approximation of real-world (as opposed to peak) performance. Read more about the criterion at the above link.

Cray XC40

Separately, Cray announced this week at the 2022 International Supercomputing Conference in Frankfurt, Germany that its Cray XC systems are now available with the latest Intel Xeon Phi (Knights Landing) processors. The company said the new Xc systems, which feature an adaptive design that supports multiple processor and storage technologies in the same architecture, deliver a 100% operation boost over prior generations. Cray also unveiled the Sonexion 3000 Lustre storage system, which tin can deliver speeds of almost 100GB/sec in a single rack.