IBM’s ambitious Blue Waters supercomputer project apparently drowned in complex technology and a flood of unexpected costs.
IBM, which was contracted by National Center for Supercomputing Applications at the University of Illinois in Urbana-Champaign, is said to have abandoned the project as the “technology […] was more complex and required significantly increased financial and technical support by IBM beyond its original expectations.”
A press release indicated that the two parties could not find a mutual plan that would accommodate the changed environment. As a result, the NCSA is getting its money back. IBM will also get back all equipment that had been already delivered.
Blue Waters was originally planned to become NCSA’s new flagship supercomputer with a peak performance of 10 PFlops that should have been delivered by at least 300,000 IBM Power7 cores. The original core architecture promised a quad-CPU module, which holds four 8-core Power7 processors. Each processor was promised to deliver a peak performance of 256 GFlops and each module about 1 TFlops. Other specs included more than 1 PB of memory, more than 25 PB of storage, 500 PB of archival storage and more than 100 Gbps of bandwidth. The NCSA said that it has not abandoned the project, but is looking for other ways to realize its next supercomputer.
SOURCE via NCSA
Remember the K — the Fujitsu supercomputer that promised to do a whopping ten petaflops by the year 2012? Well, it hasn’t reached that threshold just yet, but according to the latest Top 500 supercomputer list, it’s still faster than any other machine on Earth.
In fact, the top-ranked beast is more powerful than the next five supercomputers combined, consumes enough electricity to power about 10,000 homes for a full year, and is capable of churning out about 8.2 quadrillion calculations per second — three times as many as what runner-up (and former number-one) Tianhe-1A can process.
Today’s announcement marks the first time since 2004 that a Japanese creation sits atop Top500.org’s rankings, but Fujitsu isn’t exactly resting on its laurels. Before deploying it next year, engineers at the Riken Advanced Institute for Computational Science are aiming to add about 100,000 cores to the K’s collection of 548,352, which would provide it with even more computational muscle, and likely spell doom for all of humanity.
SOURCE via New York Times
Roll the drum boys and girls. We have yet another winner from the East. They might not have made it in the FIFA World Cup, but they’ve the world’s fastest supercomputer yo! China just passed the US and the world with their latest Tianhe-1A, fully operational at the National Supercomputer Center in Tianjin, scored 2.507 petaflops as measured by the LINPACK benchmark. That moves it past Cray’s 2.3 petaflops Jaguar located at Oak Ridge National Lab in Tennessee. Tianhe-1A achieved the record using 7,168 NVIDIA Tesla M2050 GPUs and 14,336 Intel Xeon CPUs consuming 4.04 megawatts. That’s still quite some steps to 10 petaflops, but we’ll see how the Tianhe-1A scores in next week’s Top500 supercomputers list.
SOURCE via Nvidia
Petaflop, what does it mean to you? Yes, it does sound a lot as it’s the next step after teraflop. Does 10 petaflops sounds a lot to you? Because Fujitsu has done it with the upcoming “K” supercomputer. It’s able to crunch through 10 quadrillion operations every time you blink your eye. Now, compare that to the current champ of processing farms, Cray’s Jaguar, which can handle only (only!) 1.75 petaflops of workload and you’ll know that we’re talking about a seminal leap in performance. Japan’s Riken Research Institute is the fortunate addressee on the crates of ultrafast SPARC64 VIIIfx processors that Fujitsu is now shipping out and the current plan is to have everything up and running by 2012. In total, there’ll be 80,000 CPUs, each possessing 8 cores running at 2.2GHz, which will be housed within 800 racks. So yes, there’ll be a machine somewhere on the Japanese isle with 640,000 processing cores at its disposal. Sadly you’ll never have the chance to play Crysis with this.
SOURCE via CNET
Meet MDGRAPE-3, the super computer system that is made of 201 units equipped with 24 of RIKEN’s MDGRAPE-3 LSI chips for molecular dynamics simulation (total of 4,808 chips), which are connected to 64 parallel servers equipped with 256 of Intel’s Xeon 5000-series cores and 37 parallel servers equipped with 74 Xeon 3.2 GHz cores.
The MDGRAPE-3 (Protein Explorer) is designed to perform molecular dynamics simulations of such phenomena as non-bonding interactions between atoms. Japan’s Institute of Physical and Chemical Research (Riken), SGI Japan and Intel announced the development of a supercomputer with a theoretical peak performance of 1 petaflops (one million billion floating point operations per second).