high-performance computing ➤
Sheffield-based Iceotope, providers of total liquid cooling, will be in attendance at booth #463. Te team, represented by Peter Hopton – Iceotope CEO and
inventor of the technology – will be taking a fully working system and demonstrating its approach to liquid cooling and heat recovery. Iceotope’s liquid-cooled servers have
the potential to halve the energy usage of supercomputing sites and will allow HPC users much greater freedom when designing, locating and building their facilities. Iceotope’s heat harvesting technology has been designed with research facilities in mind and will provide a compelling alternative to the traditional, air cooled servers presented at the show, upon which most HPC facilities are still based.
www.iceotope.com
NEC has been supplying complete solutions to the HPC market for more than 30 years, focusing on sustained performance and energy efficiency. Te new NEC SX-ACE is the latest product of the SX vector architecture. NEC is dedicated to continue this sole
remaining purely HPC-targeted product line in the future. At the same time, NEC is growing its market-share with the x86-based scalar NEC LX series. Tese two product lines can be tightly integrated, building hybrid systems to tackle even the most complicated workflow with both the vector and the scalar architecture, supplemented with own storage solutions, NEC’s parallel file system ScaTeFS and highly efficient direct liquid cooling to achieve optimal results. Te level of integration of all components will be further developed to reach higher levels of performance.
de.nec.com
With Numascale’s products you can have a large shared memory system at the price of a cluster. Te
NumaConnect Card that plugs into standard servers makes this possible. A cache coherent shared memory can be run it all under a single image untouched operating system. Applications in computing, visualisation and big data and many other fields benefit from this architecture. Te new NumaQ Scalable In-Memory
Analytics Appliance is launched and Numascale has new scaling data for popular HPC applications. We are eager to discuss how to run your application in shared a memory environment. Numascale also shows a new and unique way to run a large number of GPUs under the control of a single operating system.
www.numascale.com
38 SCIENTIFIC COMPUTING WORLD
PRACE, the Partnership for Advanced Computing in Europe, will be showcasing its globally competitive pan-European HPC Research Infrastructure that spans 25 nations and pools resources, leading-edge supercomputers as well as HPC expertise. Te six PRACE systems and their operations are provided by four PRACE members: BSC representing Spain, CINECA representing Italy, GCS representing Germany and GENCI representing France. Highlights at ISC14 will be the PRACE Open R&D access for industry and the PRACE SHAPE pilot (SME HPC Adoption Programme,) for small to medium-sized businesses, as well as the training offered by the PRACE Advanced Training Centres (PATC). Visitors will also be informed about the second edition of the PRACE Days, to be held in May 2015 in Dublin.
www.prace-ri.eu/
Q-Leap Networks (Booth #562) will present the newest version of Qlustar, its full-fledged HPC cluster operating
system. It is a Debian/Ubuntu based solution in the field providing many unique features. Qlustar is designed to scale from small
departmental clusters to multi-thousand node supercomputers. Its core is a lightweight, modular Compute/Storage/Cloud Node OS. It is a complete Linux distribution based on Debian/ Ubuntu. Easy and consistent cluster setup, keeping track of hard- and soſtware failures, cluster operation with ease, speed and control: QluMan, Qlustar’s management and monitoring framework takes care of this and more.
www.qlustar.com
Rogue Wave’s tools help with each stage of development, from prototyping through optimisation. Stop by booth 560 to see TotalView in action, and to learn how to build bulletproof
At ISC’14 Nvidia will be showcasing advances in applications and scientific discovery made possible with accelerated computing. Its motto this year is ‘Accelerating your breakthroughs’ and Nvidia will present how we are working with our partners to advance science. Te Nvidia booth (#230) will feature
the latest GPU technologies in accelerated computing, including the Tesla Kepler- based GPU Accelerators, whilst Nvidia experts will be on hand to answer any questions as part of its ‘Ask the GPU expert’ offering.
Nvidia will take part in several tracks and
sessions and our leading HPC centers partners will talk on OpenACC.
www.nvidia.co.uk/isc2014
@scwmagazine l
www.scientific-computing.com
soſtware and dig deep into your code to expose defects, vulnerabilities, and run-time issues, and to learn about open source licensing liability. TotalView is a scalable and intuitive debugger
for parallel applications written in C, C++, and Fortran. TotalView simplifies and shortens the process of developing, debugging, and optimising complex applications.
www.roguewave.com
TeSlurm workload manager is an open- source system designed to satisfy the needs of high performance computing. As of the November 2013 Top 500 computer list, Slurm was performing workload management on five of the 10 most powerful computers in the world including the number one system, Tianhe-2 with 3,120,000 computing cores. Te vision is to extend workload management functionality to address Exascale requirements and beyond in an open and collaborative fashion and to bring HPC technology to the world of Big Data. Slurm is highly configurable, it can be installed
and configured in a few minutes or with the use of optional plugins can provide the functionality needed to satisfy the needs of HPC centres.
www.slurm.schedmd.com
BMW’s new i3 electric car, which was designed using BMW’s HPC clusters located at Verne Global’s data centre in
Iceland, will be on display at Verne Global stand #571 at ISC’14. Te power consumption and associated high
cost of power are transforming the decision making process in terms of picking the right location to process HPC data. It does not make business sense to store and process all data in metropolitan areas which, in most instances, are subject to the highest power prices.
www.verneglobal.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 |
Page 53