November 07, 2011
With news of its world record-beating 10-petaflop K Computer still buzzing about the internet, Fujitsu has announced it's next-generation supercomputer architecture that will leave the K machine in the dust. The new system, called PRIMEHPC FX10, will scale up to 23.2 petaflops and is intended to be available commercially outside of Japan.
The platform is aimed at traditional capability supercomputing for big science application. According to the press release, PRIMEHPC FX10 is suitable for simulations that are associated with "drug discovery and medical research, analysis of earthquakes and weather phenomena, and development of new sources of energy."
PRIMEHPC FX10 will employ Fujitsu's newest SPARC processor, the SPARC64 IXfx, a 16-core chip that delivers 236.5 peak gigaflops. That's nearly twice as powerful as the 8-core SPARC64 VIIIfx CPUs that are used in the 10-petaflop K machine.
The PRIMEHPC platform will use the basic architecture of the K Computer, including the custom-built 6-dimensional mesh/torus Tofu interconnect. It will also keep the original design of a single CPU per node. The software environment, as well, is built on the K model, with the intent to maintain application compatibility between the two systems.
Scaled to the maximum number of racks (1024), a PRIMEHPC FX10 will contain more than 98 thousand CPUs, 6 PB of memory, and deliver 23.2 petaflops. At the other end of the spectrum, systems as small as 4 racks (90.8 teraflops) can also be constructed.
Unlike the K Computer, which was essentially a one-off machine for Japan's Ministry of Education, Culture, Sports, Science and Technology (MEXT), the PRIMEHPC is slated to be sold to customers both inside and outside of Japan. "The new system will be available worldwide," states the press release, which goes on to say that Fujitsu is aiming to sell 50 systems over the next three years.
Pricing was not disclosed.
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?