- 1 Vapor IO Brings OpenDCRE to General Availability
- 2 VMware Takes the Wraps Off vRealize Automation and vRealize Business
- 3 Microsoft Previews Hyper-V Containers for Windows Server 2016
- 4 Mirantis Led FUEL Project Gets Installed Under OpenStack Big Tent
- 5 Red Hat Enterprise Linux 7.2 Adds Security, DR Features
Fastest Supercomputer Breaks 10 Petaflop Barrier
The fastest computer on Earth keeps on getting faster.
The latest version of the Top500 List of the world's most powerful computers is now out, demonstrating once again that the pace of supercomputer performance continues to accelerate.
As it turns out the same top 10 supercomputers that led the Top500 list in June of 2011 are the same that lead the list in November.
"This is the first time since we began publishing the list back in 1993 that the top 10 systems showed no turnover," said TOP500 editor Erich Strohmaier in a statement.
While the same computers dominate the list, they haven't been standing still as deployments and performance have expanded.
Coming in at number one is the Fujitsu 'K' computer at the RIKEN Advanced Institute for Computational Science (AICS) in Kobe, Japan. The K Computer is the first supercomputer to break through the Petaflop barrier scoring 10.51 Petaflop/s on the Linpack benchmark. A petaflop is a unit of measure for a quadrillion calculations per second.
The K Computer is the defending supercomputer champ, and was also number one in the June 2011 ranking, though at a lower petaflop count. In June, K Computer deliver 8.16 petaflops, but has since been expanded. Fujitsu now has a staggering 705,024 SPARC64 processing cores powering the K Computer installation.
While the K Computer is the most powerful, it is also among the most efficient. The K Computer operates at an efficiency of 830 Megaflops per watt. In contrast, the average power efficiency of the Top500 supercomputers is only 230 Megaflops per watt.
The number two system in the world is the Tianhe-1A operating in the National Supercomputing Center in Tianjin China. Tianhe-1A provides 2.57 petaflops and leverages 186,368 processor cores.
Coming in at number three is the top supercomputer in the U.S, the Cray XT5 Jaguar running at the Oak Ridge National Laboratory. Jaguar scored 1.76 petaflops running on 224, 162 processor cores.
On an overall basis, the trend toward using GPUs in supercomputers has expanded in the last six months. In June only 17 systems used GPUs, while there are now 39 systems that use GPUs. NVIDIA dominates the GPU supercomputer landscape with 35 deployments.
On the CPU side Intel dominates, providing 76.8 of the processors used in the TOP500 supercomputer list.
As proof point that the world's supercomputers collectively are getting faster, the combined performance of all TOP500 supercomputers is not rated at 74.2 Petaflops/s up from 43.7 Petaflop/s in November of 2010.