1No. 1: Cray Titan
2No. 2: IBM Sequoia
3No. 3: K Computer
Housed at the RIKEN Advanced Institute for Computational Science in Kobe, Japan, the K Computer was the world’s fastest until Sequoia displaced it. The Fujitsu-based system—powered by 705,024 SPARC64 processing cores and the first to break the 10-petaflops barrier—reached a performance of 10.51 petaflops.
4No. 4: Mira
5No. 5: JuQueen
The third BlueGene/Q supercomputer from IBM to be in the top five, JuQueen is installed at the Forschungszentrum Juelich supercomputing center in Germany. The system had been ranked eighth on the list in June, but has since upped the number of Power processing cores to 393,216 and the performance to 4.14 petaflops.
6No. 6: SuperMUC
7No. 7: Stampede
This supercomputer, which is being installed at TACC, is a Dell system running the company’s PowerEdge C800 servers that are powered by Intel’s Xeon eight-core Xeon E5-2680 chips. Currently, the supercomputer, which also leverages Intel’s new Xeon Phi coprocessors, has hit a performance of 2.6 flops. However, when it’s fully built out, TACC officials expect its performance to surpass 10 petaflops.
8No. 8: Tianhe-1A
Developed by the Chinese National University of Defense Technology and installed at the National Supercomputing Center in Tianjin, China, Tianhe at one time was the fastest supercomputer in the world. It is powered by 186,368 Intel Xeon X5670 cores and more than 7,000 GPUs from Nvidia, and hit a performance of 2.56 petaflops.
9No. 9: Fermi
10No. 10: DARPA Trial Subset
New to the list, the Defense Advanced Research Projects Agency (DARPA) Trial Subset features IBM’s Power 775 servers and includes 63,360 Power processing cores. It hit a performance of 1.5 petaflops, and includes a custom interconnect from IBM.