This Week in HPC News
We’re busy preparing for ISC14 in Germany this week and have been steadily ingesting the download of pre-announcements to report as the event kicks off. Stay tuned this coming week for the breaking Top 500 announcement, coverage of new HPC system, hardware and software news, and insights from the community. We’re looking forward to it and hope to see you there.
In advance of the event, there have been several items that vendors and HPC centers didn’t hold onto for the busy news week of ISC. We’ll dive right in for this Friday round-up, catch our breath over the weekend, and prepare for the supercomputing news onslaught we see coming for International Supercomputing Week.
Los Alamos National Laboratory recently installed a new Cray system, called Wolf, which will be used for unclassified research. The 616 node machine is outfitted with two 8-core 2.6 GHz Intel “Sandybridge” processors per node, 64 GB of memory and Infiniband network. It utilizes the Laboratory’s existing Panasas parallel file system as well as a new one based on Lustre technology. The Wolf computing system operates at 197 teraflops per second.
Lawrence Berkeley National Laboratory has been named an Intel Parallel Computing Center (IPCC), a collaboration with Intel aimed at adapting existing scientific applications to run on future supercomputers built with manycore processors. The Berkeley Lab IPCC will be led by Nick Wright of the National Energy Research Scientific Computing Center (NERSC), and Bert de Jong and Hans Johansen of the Computational Research Division (CRD).
Inspur Has Big Plans – As the Chinese company stated in a release this week, they’re working on a new program, called the Inspur HPC 121 program, which is “a comprehensive market exploration project with the direct goal to take over the IBM x86 server market share. In this project, Inspur will take the emerging markets as the entry points and motivate global HPC users and partners to cooperate with Inspur. The project is believed to be a success, according to Inspur’s abilities in R&D of the fastest supercomputers, Tianhe2 and Tianhe-1A, and on the high-quality support for HPC applications and services.” We’ll certainly be following with the company at ISC to gather more details about this bold effort.
The Agency for Science, Technology and Research (A*STAR) and the Singapore Management University (SMU) will establish a Centre for Technology and Social-Behavioral Insights (CTSBI) to tap on high performance computing technology, big data analytics and behavioral sciences to study people-centric issues and human behavior. The Centre willbe officially operational July 2014, is funded by A*STAR’s Joint Council Office (JCO), which supports R&D that brings together multidisciplinary capabilities. SMU will also contribute manpower and other research resources. The Centre will be helmed by two co-directors, Dr Victor Tong from A*STAR and Professor David Chan from SMU.
The Environmental Molecular Science Laboratory (EMSL), a national scientific user facility located at Pacific Northwest National Laboratory (PNNL), has implemented DataDirect Networks (DDN) high performance block and file storage solutions to provide a storage foundation for ongoing alternative energy research projects. Following a smooth migration to its new Atipa supercomputer and DDN storage system, PNNL’s #13 ranked Cascade supercomputer delivers dramatic improvements in compute and storage power, handling up to 3.4 quadrillion calculations per second and 2.7 petabytes of usable storage.
Spectra Logic has announced that the University of Edinburgh’s Parallel Computing Centre (EPCC), has selected the company’s T380 tape library to archive a 1367 Tera-FLOPs, Cray XC30 system, the fastest supercomputer in the UK. This deployment is in support of the ARCHER (Advanced Research Computing High End Resource) project, the newest UK National Supercomputing Service and the nation’s leading high-performance computing service for academic research.
Solarflare has announced the general availability of the Solarflare Flareon Ultra SFN7142Q dual-port 40 Gigabit Ethernet (40GbE) server adapter. OpenOnload application acceleration middleware is key to the news, as it offers a chance to bypass kernel and networking overheads. In addition, with the AppFlex technology, multiple network services can be selectively delivered and controlled at the server adapter to accelerate, monitor, and secure network data. In addition to OpenOnload enablement, optional AppFlex licenses can be ordered for the SFN7142Q, including packet capture and filtering, network monitoring, and precision time synchronization (IEEE 1588 PTP) with hardware time stamping.
Univa Grid Engine has been certified and integrated with Schlumberger’s ECLIPSE industry-reference reservoir simulator software version 2014.1. This integration will provide users with an easy-to-use platform to use Univa’s workload manager in conjunction with the ECLIPSE reservoir simulator. ECLIPSE software is a popular reservoir simulation software solution that provides numerical solutions for fast and accurate prediction of dynamic behavior for all types of reservoirs and degrees of complexity including structure, geology, fluids, and development schemes.
That wraps it up for this week—we’ll see many of you and Germany and for those who aren’t making the journey, we’ll look forward to keeping you in the loop as the event and news happens.