Products HPC Products Cray XK6
Cray has announced the launch of its production hybrid supercomputing system – the Cray XK6 supercomputer – at the 2011 Cray User Group (CUG) meeting in Fairbanks, Alaska. The new supercomputer will combine Cray’s Gemini interconnect, AMD Opteron 6200 Series processors (code-named ‘Interlagos’) and Nvidia Tesla 20-Series GPUs to create a tightly-integrated system upgradeable to more than 50 Pfl ops of compute power. The Cray XK6 system offers a
unifi ed x86/GPU programming environment that includes powerful tools, libraries, compilers and third- party software. When combined with the Cray Linux Environment,
ServMax H Class The next-generation Amax ServMax H Class four- and eight-way HPC server solutions are fully optimised for the latest Intel Xeon E7 family processors, the company has announced. The H Class server solutions include extreme core counts and large memory support, providing performance and power-effi ciency across a variety of HPC deployments, including oil and gas, biomedical, research and government. The ServMax H Class platforms deliver 2.67GHz performance and
Flash 4-alpha
The University of Chicago’s Flash Center for Computational Science has released a new version of supercomputer code, called Flash 4-alpha. Based on previous software for simulating exploding stars, this is the fi rst version of the Flash code that has extensive capabilities for simulating high-energy density physics experiments. The US Department of Energy’s National Nuclear Security Administration Advanced Simulation and Computing Program has funded the addition of the new capabilities to this software, which will help scientists at universities across America better
18 SCIENTIFIC COMPUTING WORLD
understand the fundamental properties of matter at high densities and high temperatures. ‘The enhanced Flash code is an open toolset for designing and analysing experiments that address questions about the nature of planetary interiors, the creation of elements via nuclear processes, and how matter behaves in violent shocks and other extreme conditions,’ said Don Lamb, Flash Center director. The new software is a
collaboration between the Center, the Computation Institute and Argonne National Laboratory. fl
ash.uchicago.edu
TotalView Software within SGI Development Suite reproduce bugs, memory utilisation problems, and race conditions. SGI Development Suite provides an environment for developing, debugging, and performance analysis of high-performance technical computing applications for the Linux operating system. It consists of: SGI Performance Suite technical computing libraries and tools; Intel Parallel Studio XE high-performance compilers, libraries and analysis tools; and Rogue Wave TotalView Team advanced code and memory analysis and debugging tools.
www.roguewave.com
The TotalView Team application verifi cation, optimisation and debugging software from Rogue Wave Software is available as part of the SGI Development Suite software development environment. TotalView Team is used by developers tasked with developing complex applications for supercomputers. The package includes TotalView, MemoryScape, and ThreadSpotter – with ReplayEngine and Cuda support as available options. The suite provides developers with a combination of tools to pinpoint and fi x hard to
www.scientific-computing.com
the result is a fully-integrated supercomputer that blends scalable hardware, software and a network. By combining AMD Opteron processors with Nvidia GPUs, the compute node in the XK6 system gives users the option to run applications with either scalar or accelerator components. The XK6 compute blades can also be blended with XE6 compute blades into a single, multi- purpose supercomputing system. Upgradeable from the company’s XT4, XT5, XT6 or XE6 systems, the Cray XK6 system is expected to be available in late 2011.
www.cray.com
up to 40 per cent performance improvement for data-intensive software applications such as Ansys, Amber 11, Jacket and 3ds Max. The H Class features up to 10 processing cores and 20 threads, eight sockets, memory support for 2TB with Millbrook2, and additional 30MB of L3 cache. This provides the capability to address larger, increasingly complex and fl oating point applications.
www.amax.com
Platform RTM 8 and Platform Analytics 8
The latest versions of two tools designed to help HPC administrators and IT managers monitor and report on HPC workloads and analyse the associated data have been released by Platform Computing. Platform RTM 8 is a comprehensive operational dashboard that provides administrators with the information and tools needed to maximise
Altus 1750
Penguin Computing has announced the immediate availability of the new Altus 1750 server, a dual socket 1U system built specifi cally to support the fastest clock speeds available for AMD Opteron x86 CPUs. The Altus 1750 is designed to deliver the best price/
DDT version 3.0
Allinea Software has released a debugger for parallel computing on systems with hundreds of thousands of processor cores that deliver Pfl ops performance. This new version shortens debugging time and improves effi ciency for small and mid-scale users; however the new technology it incorporates also makes debugging possible for the fi rst time for users of the world’s largest supercomputers as they approach
petascale performance.
Allinea DDT 3.0 is based on a new tree-based architecture that provides a logarithmic improvement in both scalability and response time. The enhanced user interface makes it easy to debug from small to very large scales, and large-scale users will be able to manage programs that can run on supercomputers.
www.allinea.com
performance in the market for high clock speed systems, and is the only production system designed to do this for an AMD Opteron platform. It is suited to high frequency trading and other low latency applications.
www.penguincomputing.com
cluster effi ciency, enable better user productivity and contain or reduce costs. Geared toward IT managers, planners and administrators, Platform Analytics 8 is an advanced analysis and visualisation tool that enables users to identify and troubleshoot bottlenecks and analyse usage trends within the HPC datacentre.
www.platform.com
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52