GPU-accelerated cluster at Frankfurt University sets new standard for Power Usage Effectiveness (PUE)
Amsterdam, The Netherlands — ClusterVision, specialist in high performance compute, storage and database clusters, announced that its installation at the Goethe University Frankfurt has been ranked amongst the world's fastest and greenest supercomputers. In the latest release of the TOP500 supercomputers, University of Frankfurt's LOEWE high-performance compute cluster (LOEWE-CSC) is listed as the 22nd fastest in the world. It is also the fastest ATI GPU based system in the world and the fastest x86-based system in Germany.
LOEWE-CSC also ranked 8th in the Green500, a comparison of energy efficiency based on the same Linpack benchmark performance as the TOP500.
ClusterVision built LOEWE-CSC with 20,880 AMD Opteron™ 6172 processor cores and 772 ATI Radeon™ HD 5870 GPUs, using Supermicro servers and Mellanox Infiniband. ClusterVision used Bright Cluster Manager™ in both the deployment stage and as the management solution for the supercomputer.
"The LOEWE-CSC system comprises a number of cutting-edge technologies and efficiency innovations — a unique combination of computing power and pioneering work in cooling technologies," said Professor Volker Lindenstruth, Chair for HPC Architecture at the Frankfurt Institute for Advanced Studies. "The experts from ClusterVision did an outstanding job in bringing together a complex mix of advanced server, interconnect and GPU technologies. They overcame numerous unanticipated challenges to bring our unique system online, on time and exceeding our performance target. I am impressed."
"This unique architecture is the world's fastest system based on AMD/ATI GPUs, and the world's fourth fastest GPU-accelerated system," said Christopher Huggins, Sales and Marketing Director at ClusterVision. "The system reaches 60.7% efficiency, far ahead of the other leaders in the TOP500 list."
LOEWE-CSC running at less than full capacity achieved 285.2 TeraFLOPS Linpack at 736 MegaFLOPS/Watt and a 1.1 Power Usage Effectiveness ratio (PUE). PUE, a metric created by the GreenGrid, is determined by dividing the amount of power entering a data center by the power used to run the computer infrastructure within it. PUE is expressed as a ratio, with overall efficiency improving as the quotient decreases toward 1. According to the Uptime Institute, a typical data center has an average PUE of 2.5. Uptime estimates most facilities could achieve 1.6 PUE using the most efficient equipment and best practices.
"Clearly, LOEWE-CSC has blown away efficiency expectations established in the PUE metric — an impressive feat with which we at ClusterVision are proud to be associated," added Huggins. "This system is further validation of ClusterVision as a provider of many of the world's most powerful clusters. We have had over 20 entries in the TOP500 over the past 10 years, including 2 in the TOP25."
Professor Dr. Volker Lindenstruth officially launched the LOEWE-CSC supercomputer at a ceremony in Frankfurt, Germany last week.
"Professor Dr. Lindenstruth demonstrated tremendous vision and timing in undertaking this innovative approach to supercomputing," said Alex Ninaber, Founder and Technical Director at ClusterVision. "The impressive benchmark and PUE results can be directly linked to his acute attention to detail, and his tireless personal involvement, from the design of the cluster to its cooling infrastructure. Further, his whole team at FIAS should be congratulated for their ground-breaking work in developing an extremely efficient Linpack port for ATI GPUs."
About Goethe University Frankfurt
The Goethe University's Hessian high performance computer organisation (HHLR-GU) coordinates all IT and HPC activities at the Goethe University. Part of HHLR-GU is the Center for Scientific Computing, which was founded as a joint initiative of research groups from the university departments Physics, Chemistry, Biochemistry and Pharmacy, Geosciences, Computer Science and Mathematics, and the Frankfurt Institute for Advanced Studies (FIAS). The mission of the CSC is HPC support for the scientific community and education in Computational Science. The CSC organizes the interdisciplinary master program Computational Science and regular seminars. It will also operate the new flagship computer LOEWE-CSC, built by ClusterVision.
ClusterVision is specialist in the design, implementation and support of small- and large-scale computer clusters. Their clustering technology provides a cost-effective alternative to traditional supercomputers by connecting multiple computers to form a unified powerful computing system. ClusterVision's team of experts has designed and built some of the largest and most complex computational, storage and database clusters in Europe. With a background in applied scientific research and practical experience with a wide range of HPC technologies, the team understands customers' requirements and provides tailor-made solutions. ClusterVision has offices in Amsterdam, Gloucester (UK), Munich, Paris, Geneva, Milan, Oslo and Madrid.
For more information
1045 AL Amsterdam
Tel: +31 20 407 7550
Fax: +31 84 759 8389
christopher [dot] huggins [at] clustervision [dot] com
|The new supercomputer at the University of Frankfurt ranks 8th in the Green500.
|The new supercomputer at the University of Frankfurt runs Bright Cluster Manager — Advanced Edition