SEATTLE, Wash.—Fujitsu’s massive K Computer remains the world’s most powerful supercomputer, being the first system to break through the 10-petaflop (quadrillions of floating point operations per second) barrier. The latest Top500 list of the world’s fastest computers was released Nov. 14 on the opening day of the SC 11 supercomputer show here. The K Computer, which is being installed at the Riken Advanced Institute for Computational Science in Kobe, Japan, and is powered by Fujitsu’s latest SPARC64 processors, hit a peak performance of 10.51 petaflops. That the system runs on Fujitsu chips is somewhat unusual: Intel is the top chip vendor on the Top500 list, powering 76.8 percent—or 384—of the supercomputers on the list. Second on the list is fellow x86 chip maker AMD, which powers 63 systems, or 12 percent of the servers on the list. In addition, the K Computer also bucks a growing trend using GPU accelerators from the likes of Nvidia and AMD in hybrid systems to increase computational speed and increasing energy efficiency. Here, eWEEK looks at the world’s Top 10 most powerful supercomputers.
of
K Computer
The supercomputer is powered by 705,024 of Fujitsu’s SPARC64 processors spread over 864 server racks. When the system entered the list at No. 1 in June, it marked the first time since 2004, when Fujitsu’s Earth Simulator was knocked off the perch, that a Japanese computer headed the list.
Tianhe-A1
Developed by the Chinese National University of Defense Technology, the Tianhe-A1 was knocked off the top of the list in June by the K Computer. Powered by more than 14,000 Xeon X5670 chips and more than 7,000 GPUs from Nvidia, the system, installed at the National Supercomputing Center in Tianjin, has hit a peak of 2.5 petaflops.
Jaguar
Installed at the Oak Ridge National Laboratory in Tennessee, the Jaguar supercomputer is powered by Cray’s XT5-HE technology powered by AMD’s six-core Opteron chips. The system, which has hit a peak of about 1.75 petaflops, is about to be revamped by Cray using AMD’s new 16-core Opteron 6200 chips and GPUs from Nvidia. Once completed, the new system—dubbed Titan—will have theoretical peaks of 10 to 20 petaflops.
Nebulae
Built from Dawning’s TC3600 blades using Intel’s six-core Xeon 5650 and Nvidia GPUs, the Nebulae at the National Supercomputing Center at Shenzhen has hit 1.27 petaflops.
Tsubame 2.0
Installed at the GSIC Center at the Tokyo Institute of Technology, the Tsubame 2.0 comprises HP’s ProLiant SL390 G7 systems, which are powered by Intel’s six-core Xeon 5670 processors and leverage Nvidia GPUs. The system, which runs both Linux and Windows, has a performance of 1.19 petaflops.
Cielo
Cielo was built by Cray using its XE6 systems at the Los Alamos National Laboratory for the National Nuclear Security Administration. Cielo leverages AMD’s eight-core Opteron 6136 chips for a performance level of 1.11 petaflops.
Pleiades
The NASA supercomputer at the Ames Research Center comprises SGI’s Altix ICE 8200EX/8400EX servers powered by Intel’s Xeon HT QC 3.0, 5570 and 5670 processors. It offers a performance of 1.08 petaflops.
Hopper
Another Cray supercomputer, this one for the National Energy Research Scientific Computing Center at the Lawrence Berkeley National Laboratory. It leverages Cray’s XE6 systems, which are powered by AMD’s 12-core Opteron 6172 processors. The system, with a total of 153,408 cores, offers a performance of 1.05 petaflops.
Tera-100
Housed at France’s Atomic Energy Commission, the Tera-100 last year became Europe’s first supercomputer to break the petaflop barrier. Built by Bull, the system is a cluster of 4,370 bullx S series servers, with more than 17,400 Xeon 7500 chips from Intel. Its performance is listed at 1.05 petaflops.
Roadrunner
The IBM-built supercomputer, installed at the Los Alamos National Laboratory, is a cluster of BladeCenter systems powered by Big Blue’s 3.2GHz PowerXCell 8i and AMD’s dual-core Opteron chips. Its performance is listed at 1.04 petaflops.
Windows Azure is a public cloud platform for building, hosting and scaling applications. Try Windows Azure free for 90 days and get 20GB outbound and unlimited inbound data transfer.
IT & Network Infrastructure News & Reviews 
Cloud Data Backup Now Makes Sense for SMBs: 10 Reasons Why 
HP, Lenovo, Asus Laptops, Desktops Offer Consumer Style for Busin[..]
Microsoft`s F#: 10 Reasons Why It`s a Hot Programming Language fo[..]
Data Center Efficiency: 10 Tips to Boost Productivity, Reduce Pow[..]
iPhone Apps to Kick Off the 2012 Summer Season 
Email Security: 10 Steps for Dealing With Dangerous Messages
See All Slideshows
SEATTLE, Wash.—Fujitsu’s massive K Computer remains the world’s most powerful supercomputer, being the first system to break through the 10-petaflop (quadrillions of floating point operations per second) barrier. The latest Top500 list of the world’s fastest computers was released Nov. 14 on the opening day of the SC 11 supercomputer show here. The K Computer, which is being installed at the Riken Advanced Institute for Computational Science in Kobe, Japan, and is powered by Fujitsu’s latest SPARC64 processors, hit a peak performance of 10.51 petaflops. That the system runs on Fujitsu chips is somewhat unusual: Intel is the top chip vendor on the Top500 list, powering 76.8 percent—or 384—of the supercomputers on the list. Second on the list is fellow x86 chip maker AMD, which powers 63 systems, or 12 percent of the servers on the list. In addition, the K Computer also bucks a growing trend using GPU accelerators from the likes of Nvidia and AMD in hybrid systems to increase computational speed and increasing energy efficiency. Here, eWEEK looks at the world’s Top 10 most powerful supercomputers.