August 24, 2012
This summer’s TOP500 list could be described as a changing of the guard with surprising countries capturing top 10 spots and the US reclaiming first place. Although China was the only BRIC nation (Brazil Russia, India, and China) to have any top ten systems, Russia is certainly making an effort to join that elite club.
Their Lomonosov supercomputer is currently ranked 22nd, with a Linpack rating of 901.9 teraflops. The country is keeping their sights on the top of the list though, with plans to deploy a faster system in the near future.
Today, Russia Beyond the Headlines reported on an announcement made by the Russian Academy of Sciences (RAS). The Academy plans to build a petascale supercomputer that rivals some of today’s fastest systems. The initial implementation is expected to come online by year’s end with a capacity of 2 petaflops. As such, the new supercomputer would likely become the country’s fastest system and most likely capture a top 10 ranking.
Reaching the petascale benchmark is certainly a feat upon itself, but the RAS has further plans for the unnamed system. After upgrades, the final deployment is expected to have a capacity of 10 petaflops. Currently, only two systems in the world can make that claim, Sequoia at Lawrence Livermore National Laboratory and Japan’s K computer.
In the Russia Beyond the Headlines report, Boris Shabanov, deputy director of the Joint Supercomputer Center at RAS, described the deployment timeline. “Our goal is to complete the first two elements by the end of the year, and then the technology will make it possible to upgrade its capacity to 10 petaflops within a reasonable time – over the next year – provided there is adequate investment,” he said. The RAS system is expected to cost around $63 million.
If the machine is fully upgraded over the next year as Shabanov expects, it may claim a top five position. By at that point, it will be competing with even more powerful systems, like Titan at Oak Ridge National Laboratory and Blue Waters at NCSA, as well as possible new entries from China.
Not much has been revealed regarding the specifics of the system, but Shabanov noted that energy efficiency is a key feature of the upcoming super and will employ a “unique cooling system” as well as and cutting-edge x86 coprocessors. The suggests the machine will rely on Xeon Phi (MIC) chips, Intel’s manycore acceleratore that is expected to debut before the end of the year. CPU and interconnect details weren’t revealed.
Russia began competing more seriously in the supercomputing arena after President Medvedev tasked scientists to catch up with the West. Over $140 million went into HPC in 2011, and over a billion dollars is being allocated to support Russia’s exascale push.
Full story at Russia Beyond the Headlines
10/30/2013 | Cray, DDN, Mellanox, NetApp, ScaleMP, Supermicro, Xyratex | Creating data is easy… the challenge is getting it to the right place to make use of it. This paper discusses fresh solutions that can directly increase I/O efficiency, and the applications of these solutions to current, and new technology infrastructures.
10/01/2013 | IBM | A new trend is developing in the HPC space that is also affecting enterprise computing productivity with the arrival of “ultra-dense” hyper-scale servers.
Ken Claffey, SVP and General Manager at Xyratex, presents ClusterStor at the Vendor Showdown at ISC13 in Leipzig, Germany.
Join HPCwire Editor Nicole Hemsoth and Dr. David Bader from Georgia Tech as they take center stage on opening night at Atlanta's first Big Data Kick Off Week, filmed in front of a live audience. Nicole and David look at the evolution of HPC, today's big data challenges, discuss real world solutions, and reveal their predictions. Exactly what does the future holds for HPC?