ISC’11 show preview
PCIe links. Now each 1u server can access from one to four Tesla M2090 GPUs without burdening the server with the cost overhead of GPUs, extra power and cooling.
www.nextio.com
business analytics applications. Director of product management, Bill McMillan, will be speaking at the Exhibitor Forum during the show, addressing the question: ‘Scalability: what does this really mean?’ The company will be exhibiting at booth 230.
www.platform.com
Using AMD’s standardised HyperTransport processor-memory interface, Numascale is the fi rst company to develop and commercialise an interconnect with the ability to build scalable high-performance clusters with cache coherent shared memory. The NumaConnect SMP adapters enable the whole cluster to be utilised as a single system running one instance of the operating system. Standard applications can take advantage of all available processors, memory and I/O throughout the cluster with no software overhead or system memory loss. The Numascale team will be demonstrating the technology at booth 710.
www.numascale.com
ParStream addresses the problems arising from rapidly increasing data volumes, such as web analytics and scientifi c application scenarios, by employing algorithms and patented index structures. ParStream has been developed using parallel architectures and modern hybrid CPU/GPU systems. It is the fi rst database engine designed and optimised for processing extreme amounts of data on highly parallel processor architectures, and is a scalable solution suitable for any server infrastructure.
www.parstream.com
PBS Works, a division of Altair Engineering, is showcasing its HPC solutions in workload management, job scheduling, and analytics at ISC’11. This year marks the 20th anniversary for PBS Professional, the workload management solution for HPC, which optimises the scheduling and management of complex workloads, enabling maximum, effi cient use of computing resources. Scaling to support hundreds of thousands of processors, it is the only workload management solution to be EAL3+ security certifi ed. Live demonstrations of the PBS Works
www.scientific-computing.com
Nvidia will be presenting its Tesla technology. Based on the Cuda architecture codenamed ‘Fermi’, the Tesla M-class GPU modules are parallel processors for high- performance computing. Tesla GPUs are particularly suited for seismic processing, biochemistry simulations, weather and climate modelling, signal processing, CAE, CFD, and data analytics. The Tesla 20-series GPU processors
are the fi rst to deliver greater than 10 times the double precision horsepower of a quad-core x86 CPU. Based on the Fermi architecture, these GPUs feature up to 665 Gfl ops of double-precision performance, one Tfl op of single- precision performance, ECC memory error protection, and L1 and L2 caches. The Tesla M-class GPU modules are
integrated into GPU-CPU servers from OEMs, offering greater choice in how GPUs are deployed within datacenters. A variety of rackmount and blade systems, and remote monitoring and management capabilities, enable large scale-out deployments.
www.nvidia.eu/tesla
suite, highlighting energy savings with Green Provisioning, end-user productivity with Cloud Computing, and improved application performance with topology- aware and GPU-aware scheduling, will be taking place at booth 820.
www.pbsworks.com
Platform Computing is a long-standing sponsor of the International Supercomputing
Conference. The company recently announced the next generation of its high-performance computing (HPC) management product, Platform HPC 3, as well as introduced Platform RTM 8 and Platform Analytics 8, designed to monitor, manage and analyse cluster usage. In addition, the company announced support for the Apache Hadoop MapReduce programming model to bring enterprise-class distributed computing to
Exhibiting at booth 125, PRACE (the Partnership for Advanced Computing in Europe) project members will present the current and future steps of PRACE. The organisation has a BoF on Tuesday 21 June at 2.30pm to 3.15pm in Hall B. The title of this BoF is ‘E-infrastructures for science in Europe’ (BoF #1 in the ISC’11 agenda). Appearing for the third time, the PRACE Award will be presented at the opening session of the show. A treasure hunt is also being organised and attendees are invited to attend the booth for further details.
www.prace-ri.eu
Rogue Wave Software is an independent provider of cross-platform software development tools and embedded components for the next generation of high-performance computing applications. The full-day tutorial, ‘Hands-On Debugging and Optimizing Parallel Programs,’ will take place on Sunday 19 June. The company will be at booth 632, demoing TotalView with Cuda debugging, reverse debugging with ReplayEngine, and performance analysis with ThreadSpotter, pinpointing issues in memory bandwidth and latency, data locality and thread interaction.
www.roguewave.com
RSC, a Russian provider of complete HPC solutions, will be at booth 840 showcasing energy-effi cient technologies implemented in new and fl exible HPC systems that include high-speed interconnects and the company’s own cluster software. Complete system optimisation approaches PUE<1.2, with electricity savings of more than 60 per cent compared to traditional HPC clusters. The RSC system, based on the latest Xeon processors and commodity components, offers fl exibility and interoperability, while advanced liquid cooling results in high compute density among x86 clusters. Standard racks deliver more than 40 Tfl ops with Sandy Bridge microarchitecture, and improve the use of datacenter space by up to three times compared to air-cooled systems. The vibration-free design and effi cient cooling make it suitable for long-running scientifi c and engineering applications.
www.rscgroup.ru
JUNE/JULY 2011 29
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