(Auszug aus der Pressemitteilung)
NEW ORLEANS, LA — Nov. 15, 2010 – At Supercomputing 2010 (SC10), AMD today announces that 24 of the top 50 fastest supercomputers ranked in the bi-annual TOP500 list are using AMD’s CPU and/or GPU technology, including half of the top 10. 59 supercomputers in the TOP500 list are now taking advantage of the massive compute power offered by AMD Opteron™ platforms, including 22 systems based on the new 8- and 12-core AMD Opteron 6000 Series platform.
SC10 marks yet another significant milestone for the HPC community, as the heterogeneous computing trend which AMD has helped pioneer is driven further into mainstream HPC deployments, helping enable incredible advances in science, research and academia. AMD continues to drive this trend forward; delivering world leading performance and efficiency with the LOEWE high-performance computer cluster at Goethe University of Frankfurt based on 12-core AMD Opteron 6100 Series processors and ATI Radeon™ HD 5870 GPUs, achieving top rankings in the TOP500 and Green500 lists. This development builds on AMD’s rich history of innovative technology, strategic partnerships, and industry-leading OEM customers who creatively use AMD platforms to consistently drive improved performance.
“AMD has led the HPC community beyond the petaflop barrier and we believe heterogeneous computing will emerge as the favored formula for driving HPC performance to the next level,” said Pat Patla, vice president and general manager, Server and Embedded Division, AMD. “AMD pioneered 64-bit computing and native multi-core processing, and with world-class CPU and GPU technology AMD is ready to lead the heterogeneous computing era.”
AMD Opteron processors and AMD GPU technology have historically powered supercomputers in the world’s largest and most prestigious national labs, research institutions and private enterprises. With TOP500-ranked supercomputing systems in 11 countries around the world, AMD’s global footprint currently includes these world-class systems:
- #2: Jaguar, a 1.75 petaflop Cray XT5™ supercomputer based on the Six-Core AMD Opteron processor and having almost a quarter of a million cores
- #5: Franklin, a 1.05 petaflop Cray XE6™ based on the 12-core AMD Opteron processor
- #7: Roadrunner, a hybrid system from IBM using AMD Opteron processors in conjunction with IBM Cell technology located at Los Alamos National Laboratory
- #8: Kraken, the fastest academic supercomputer in the United States from the University of Tennessee harnessing Six-Core AMD Opteron HE processors
- #10: Cielo, a joint Cray XE6 supercomputer from Los Alamos National Laboratories and Sandia National Laboratories leveraging 8-core AMD Opteron 6100 series processors
During SC10, AMD will show demos in booth 3119 with partners including Appro, Cray, Dell, HP, IBM and Supermicro.