Celebrating its 25th year, SC is expected to attract more than 350 exhibitors spread over nearly 140,000 square feet of exhibit space.
Adaptive Computing (booth #3113) will be showing its Moab optimisation and scheduling software suites that power many of the world’s largest supercomputing environments and many systems on the HPC Top500 list. Moab enables large enterprises in oil and gas, financial services, manufacturing, research and government to perform computationally intensive simulations and analyse big data faster, more accurately, and more cost effectively.
Moab gives organisations a competitive advantage, inspiring them to develop cancer-curing treatments, discover the origins of the universe, lower energy prices, manufacture better products, improve the economic landscape and pursue game-changing endeavours. With a converged HPC, data centre, and cloud environment, data analysts can speed the time to discovery. Adaptive Computing holds more than 32 patents and has pioneered many aspects of private/hybrid cloud, technical computing and large-scale scheduling.
Allinea Software (#1719) is a trusted leader in software development tools for parallel programming and high-performance computing. At SC13, attendees can discover how HPC centres and users are reducing development times and increasing application performance with Allinea Software's easy to use unified debugging and profiling tools. Visitors can also learn how Allinea DDT and Allinea MAP can help them to get more from Intel Xeon Phi coprocessors and Nvidia Cuda 5.5 accelerators, and enter a profiling competition which the company says will bring an exciting twist to application performance.
Allinea Software is participating in the official programme at: ‘Debugging MPI and Hybrid/Heterogeneous Applications at Scale’, SC13 tutorial session on Sunday 17 November in room 404; 2013 Workshop on Extreme-scale Performance Tools, ‘Performance Profiling and Debugging at the Extreme Scale and Beyond’, Monday 18 November in room 501 at 4.10pm; and ‘Pick Your Battles: Getting Results Faster with Intel Xeon Phi and Nvidia Cuda’, Thursday 21 November in room 501/502 at 3.30pm.
Altair's (#2513) technical cloud solutions address the unique needs of enterprises by simplifying access to HPC infrastructures, allowing engineers and scientists to focus on their work. At SC13, the company will be demonstrating PBS Professional, which optimises the world’s largest supercomputers. It is the workload management of choice for many supercomputing sites within the Top500.
Altair’s web solutions will also be on display: PBS Analytics is a portal that provides administrators with advanced job analyses to support data-driven planning; Compute Manager is a job submission portal that’s application aware, so end users only concern themselves with applications they wish to run; and Display Manager is a remote visualisation tool to view datasets for rapid collaboration.
Altair's public cloud solution, HyperWorks On-Demand, brings SaaS, PaaS and IaaS to users within a single and intuitive web-based portal, and HyperWorks Unlimited, the company’s managed, private cloud solution with fully configured hardware and software, offers users unlimited use of all Altair software within the appliance.
Asetek (#4329) is featuring RackCDU D2C and RackCDU ISAC hot water liquid cooling for HPC and data centres. On display will be liquid-cooled servers from Cray, including the CS300-L and servers retrofitted with liquid cooling from Cisco, Intel, Supermicro and others. Asetek’s booth will include a running 92 node cluster cooled by RackCDU D2C with monitoring software providing real-time reporting, alerting and the ability to integrate into DCIM software.
Asetek’s RackCDU D2C (Direct-to-Chip) provides cooling cost reductions up to 80 per cent and density increases of 2.5x-5x. RackCDU ISAC (In-Server Air Conditioning) provides cooling cost reductions exceeding 80 per cent and enables operation without concern for air quality.
Asetek, the company says, invented sealed loop, low-cost, liquid cooling systems for computers. With more than 1.5 million liquid coolers sold, Asetek offers performance, reliability and production at scale. OEM customers include Cray, Dell, HP, AMD and Intel.
As well as appearing on its own stand, Boston (#441) will be partnering with ARM (#3141), CoolIT Systems (#1916), Mellanox (#2722), Nvidia (#613) and Supermicro (#3132). At the show, Boston will be unveiling the World's first ARM-based enterprise server, the Boston Viridis, which uses Calxeda Energy cards to make pedal-powered data centres a reality. In partnership with CoolIT, Boston will also be showing some liquid cooling configurations and giving an insight into the other components that can be liquid cooled in the future.
Further highlights include a demonstration of how to get started with the ARM architecture with help from the cloud-based ARM-as-a-Service; an introduction to the latest developments in GPU computing, and how GPUs integrate with Boston solutions; demonstrations of how Intel Enterprise Edition expands high-performance computing into the commercial enterprise, which will take place on Supermicro’s stand; and how Mellanox solutions can reduce the number of under-utilised components.
In keeping with our tradition, visitors to the Bright Computing booth (#1725) at SC13 will have the opportunity to experience the present and future Bright Cluster Manager. Bright has an established reputation as the management solution of choice within leading public and private-sector organisations across the globe. These organisations depend on Bright to provision, monitor and manage their on-premise compute clusters, as well as to extend these clusters seamlessly into the cloud. The next major release of Bright allows a bare-metal cluster to be executing Hadoop workloads within an hour. From monitors, metrics and health checks, to big data workloads, Bright is the single management solution that works with most Hadoop distributions. Alongside HPC clusters, Hadoop clusters can seamlessly be extended into the cloud through Bright. Taken together, Bright eases the management of all types of clusters and clouds. Using the Bright GUI, we can demonstrate all of this during your visit.
Bull (#2313) will show its range of extreme computing solutions for every type of need. From energy efficient data centre design to expert advice on optimisation for many-core environments, including of course Bull’s Direct Liquid Cooled HPC servers, as well as power consumption optimisation features in the software environment, Bull’s offer is totally focused on extreme energy efficiency.
The bullx DLC B700 range features a direct liquid-cooling technology that makes it possible to use warm water instead of chilled water, thus boosting energy performance by around 40 per cent compared to traditional data centres, despite being just as easy to maintain as standard, air-cooled servers. The bullx DLC B700 range is based on industry-standard components, which are all cooled through direct contact with a cold plate. The bullx DLC B700 features the latest processor technologies, including also Intel Xeon Phi coprocessors and Nvidia Tesla GPUs.
Caps entreprise (#1127) will present the latest features of its flagship compiler, Caps OpenACC, and will launch a new entity, Caps engineering. Capitalising on 12 years of experience in developing parallel compilers and its expertise in processor architectures, Caps has created Caps engineering to answer its customers' requests. This new entity offers a large range of consulting services to make applications efficiently parallel, such as application diagnosis for the performance, code porting and code optimisation, as well as a set of parallel computing courses to get the relevant expertise in OpenACC, Cuda, OpenCL, OpenMP, Intel Xeon Phi Programming, etc.
Chelsio Communications (#3306) is a leading technology company focused on solving high performance networking and storage challenges for virtualised enterprise data centres, cloud service installations, and cluster computing environments. Now shipping its fourth-generation protocol acceleration technology, Chelsio is delivering hardware and software solutions including Unified Wire Ethernet network adapter cards, unified storage software, high performance storage gateways, unified management software, bypass cards, and other solutions.
At SC13, Chelsio will demonstrate, T5 ASIC based Unified Wire Network Adapter T580 and its superior iSCSI SAN performance. In addition, Chelsio will demonstrate its iWARP (RDMA over Ethernet) technology running at 40 Gb.
CoCoLink (#4509) has provided supercomputing systems and applications since 2001. Drawing on a rich experience in high-performance computing, the company has developed a supercomputing system with what it says is a unique and innovative architecture.
At SC13, CoCoLink will be showing CliC-80000, a supercomputing system which maximises performance with various computing accelerators such as GPU, MIC and FPGA. CliC-80000 supports 8~16 units of computing accelerators per node with built-in Infiniband FDR, and in a world’s first, CliC-80000 provides P2P data transmission for equipped accelerators (data transmission from each to each). This P2P function ensures that CliC-80000 is an efficient and powerful computing system. In addition, CoCoLink is going to launch a new system which supports 18 units of accelerators.
Coolcentric (#1241) delivers one of the world's most energy and space efficient heat removal solutions for data centres. The Coolcentric family of heat exchangers comprise passive, liquid-cooled heat exchangers; standard RDHx; and low Density RDHx-LD to replace standard rear doors on IT rack enclosures. Close-coupled to the IT enclosure, heat exchangers remove heat as close to the heat source as possible, providing the ultimate containment solution. Coolcentric heat exchangers are reliable, scalable, energy efficient and space-saving turn-key cooling solutions that help drive down capital and operation costs. They can provide sensible cooling for rack loads from 5kW to 40kW.
Factory certified technicians are available to provide installation, commissioning and preventive maintenance throughout the world.
CoolIT Systems (#1916) is featuring a variety of DCLC solutions, including the AHx and CHx systems in 45U, 25U and 12U arrangements, at this year’s supercomputing event. CoolIT's Direct Contact Liquid Cooling (DCLC) technology delivers what the company says is industry leading data centre solutions for the HPC, cloud and enterprise markets.
The Rack DCLC product line provides scalability, modularity and flexibility to meet even the most aggressive density, efficiency and performance requirements. With more than 40 issued patents, 12 years of product innovation and more than one million CPU’s being chilled by CoolIT’s DCLC technology, the company offers some of the most advanced direct contact liquid cooling solutions in the world.
Cycle Computing (#3610) invites attendees of SC13 to test their knowledge of supercomputing history and get a preview of the next era in HPC by visiting its booth. As a leader in utility HPC software, Cycle Computing enables applications to run between computing resources, and over the Cloud. From 50 to 50,000+ cores and 100s of TBs of data, the world’s brightest minds rely on Cycle Computing software to tackle their most challenging problems in less time and a lower cost than ever before.
Cycle Computing is the single pane of glass from which customers and partners easily orchestrate complex workloads and data across internal and external resources. An entirely privately held, profitable company, Cycle Computing’s clients include Fortune 500s, SMBs, government and academic institutions including JP Morgan Chase, Pfizer, and Lockheed Martin.
Visit Booth #4213 to learn more about GÉANT’s extensive global connectivity (reaching over 100 countries worldwide) and advanced services such as Bandwidth on Demand - as well as our role in global networking standards. Find out more - http://www.geant.net/MediaCentreEvents/Events/Pages/GEANT-at-Super-Computing-2013.aspx
DataDirect Networks (DDN) (#1307) is a leader in HPC storage. Powering more than two thirds of the world’s fastest supercomputers, DDN is uniquely positioned to deliver solutions for the emerging data-centric computing era. Visitors to the DDN booth will be able to see how the company’s technologies enable data-intensive environments to integrate, streamline and accelerate computing workflows, while significantly reducing infrastructure cost and complexity.
DDN will preview burst buffer concepts and demonstrate current products spanning block, file and object storage including application-aware hybrid storage, integrated parallel file system appliances, accelerated analytics, and secure private cloud.
Register for DDN’s annual user group at www.ddn.com to see how leading HPC sites around the world are tackling petascale big data problems today and preparing for exascale. And also see Dr James Coomer’s presentation on real-world results of implementing SSDs to benefit parallel file system performance.
Founded at the University of Edinburgh in 1990, EPCC (#3932) is a leading European centre of expertise in advanced computing research, high-performance computing and technology transfer. The CRESTA project, which EPCC leads, will share the European Exascale Projects booth (#3741) with the other major European Exascale initiatives. Alongside the exhibits, EPCC is involved in a number of the workshop and Birds of a Feather (BoFs) sessions. It is organising the ‘Building on the European Exascale Approach’, ‘HPC Training Perspectives and Collaborations from PRACE (Europe), XSEDE (US) and RIKEN AICS (Japan)’, and ‘G8 Extreme Scale and Big Data Program: Progress on Current Exascale Projects and Future Funding Opportunities’ BoFs. Furthermore, EPCC staff are organising a workshop on ‘Exascale MPI’.
European Exascale Projects will be exhibiting on booth #3741 at SC13. The drive to develop, produce and exploit exascale platforms has led to distinct strategies across continents, differences in expertise and drivers defining the approaches taken. In Europe, significant investment has been made in three complementary research projects: CRESTA, DEEP and Mont-Blanc, with a combined funding of €25 million.
As well as developing novel prototype hardware, these projects are defined by their co-design approach to software and application development, their investment in European leadership tools and scientific applications, and their cross-project collaboration as part of a European exascale community. These projects are now well established and the booth will showcase early hardware, highlight scientific application successes and demonstrate prototype software and tools.
Finisar (#3608) is a global technology leader for fibre-optic subsystems and components that enable high-speed data, video and voice communications for high-speed computing, telecommunications, networking, storage, wireless, and cable TV applications. For 25 years, Finisar has provided critical optics technologies to system manufacturers to meet the increasing demands for network bandwidth.
At SC13 in Denver, Finisar will be showcasing its transceiver and active optical cable portfolio, which is built on Finisar’s experience in high-speed VCSEL technology. The company will showcase interconnects that transmit data at aggregate rates up to 400 Gbps, based on VCSELs operating in parallel lanes up to 25 Gbps each. Visitors to the booth can learn more about Finisar and how the company is enabling the next generation of supercomputers.
Fujitsu (#2718) is positioned at the forefront of the supercomputing space with 30 years' experience in the successful development of high-performance systems. To meet individual requirements, Fujitsu offers scalable HPC architectures from x86 Celsius high-end workstations and x86 Primergy industry-standard servers all the way up to the PrimeHPC FX10 supercomputer. Network components, Eternus storage platforms, the HPC Cluster Suite including the HPC Gateway, and consulting and integration services complete the offering.
Thoroughly designed to operational and workflow-related requirements, the company’s segment and workload-specific integrated HPC offering provides an optimal price/performance ratio. Ready-to-go delivery of HPC infrastructures guarantees rapid deployment for production and includes Intel Cluster Ready certification for proven performance and trouble-free operation. Last but not least, application integration (onboarding) makes the use of HPC easier for experts and more broadly accessible to beginners.
The German Climate Computing Centre (DKRZ) (#329) started its operation as a unique national service centre for climate research in 1989. Over the past 25 years, the available computing power and volume of archived data has grown literally a million-fold. Its mission statement is based on three major goals: maximum computing power, advanced data management, and expert service.
German climate research depends on access to the HPC systems at DKRZ. Climate modelling produces enormous amounts of data – at DKRZ more than 25 Petabytes are stored. DKRZ provides tools and services to deal with the typically large data sets.
At SC13, DKRZ will present the visualisations of intricate model experiments. These contributed to the fifth IPCC assessment report published on 27 September 2013. The extensive climate simulations were carried out by DKRZ and Max-Planck-Institute for Meteorology and occupied one quarter of computing capacity over a period of two years.
Gompute (#416) delivers comprehensive solutions for high-performance computing in-house, in-the-cloud or both. With more than 10 years' experience working with industrial clients, providing solutions and services to the IT and engineering communities, as well as advisory functions to C-level executives, it delivers a collaborative and highly productive work environment for geographically distributed engineering teams.
Combined with Gompute On-Demand, the company ensures that users have the option to either scale out or completely host their HPC environment, ranging from one to 1,000s of cores in a Gompute-owned supercomputing centre.
Green Revolution Cooling (#2625) provides data centres with liquid submersion cooling solutions that require low overhead investment, use extremely low amounts of energy and allow for incredible cost savings from day one. The CarnotJet system safely and efficiently brings organisations’ PUE down to around 1.05 or less by using 95 per cent less cooling power, 10-25 per cent less server power, and by dramatically reducing infrastructure requirements. Green Revolution Cooling’s solutions use a non-toxic, non-conductive and long-lasting mineral oil coolant known as GreenDEF Coolant, which has 1,200 times more heat capacity by volume than air and never needs to be replaced.
At SC13, see the cooling solution that is already saving forward-thinking organisations millions on power consumption in six countries across the globe. Visitors to the booth can see why Green Revolution Cooling and the CarnotJet system continue to grow in prominence both internationally and domestically as a benchmark for data centre efficiency.
At this year’s SC13, IBM (#2173) will feature a broad array of solutions that can help organisations accelerate insights and speed results with both technical sessions and space in the Exhibit Hall. In the Exhibit Hall, attendees will find solutions from IBM Technical Computing and IBM Platform Computing, as well as a display of past generations of HPC technology in celebration of the 25th anniversary of the Supercomputing Conference and Expo.
IBM is also participating in a number of technical sessions, which cover topics in research, government, finance, and big data. For example, the HPC in Finance Workshop – chaired by IBM Research, FountainHead and the University of San Francisco – brings together practitioners, researchers, vendors and scholars from the complementary fields of computational finance and HPC to promote an exchange of ideas, develop common benchmarks and methodologies, discuss future collaborations and determine new research directions.
Additionally, the IBM Research initiative X10 session will focus on a high-performance, high-productivity programming language. To find additional IBM sessions in the Technical Programme, view the programme schedule sc13.supercomputing.org/schedule.
Sheffield-based Iceotope, the next generation liquid cooling specialist (Booth 3918), will share the details of its latest customer deployment, the first Iceotope system to be located outside the UK, as well as providing attendees with an overview of its innovative cooling solution.
Iceotope will be represented by Peter Hopton, the founder and inventor of the technology, as well as Keith Deakin who has engineered that vision into reality.
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, the likes of which most HPC facilities are still based.
Mellanox Technologies (#2722) is a leading supplier of end-to-end InfiniBand and Ethernet interconnect solutions and services for servers and storage. Mellanox interconnect solutions increase data centre efficiency by providing the highest throughput and lowest latency, delivering data faster to applications and unlocking system performance capability.
Mellanox offers a choice of fast interconnect products: adapters, switches, software, cables and silicon that accelerate application runtime and maximise business results for a wide range of markets including high-performance computing, enterprise data centres, Web 2.0, cloud, storage and financial services.
Motivair (#318) will be showing its Chilled Door HPC rack cooling system at SC13. Designed to remove 100 per cent of server heat at its source, the Chilled Door is a non-invasive and simple addition to new or retrofit HPC rack applications, with no direct liquid contact. It removes >45kW from any rack, using up to 65F chilled water for condensate-free operation, while a sealed interface frame facilitates accurate fit to any standard or custom rack.
Generous, low power EC fan capacity and an overriding fan pressure switch always ensure safe operation under any operating mode. Enhanced PLC software provides remote communication via BACnet LON or Modbus. Furthermore, a unique leak prevention system sets the alarm, isolates the Chilled Door from the water supply and increases the fan speed. Any alarm condition switches the door-size blue LED to red for instant recognition.
NASA's exhibit (#882) will highlight science and engineering results enabled using the agency's world-class supercomputing resources. NASA will showcase cutting-edge work in aeronautics, Earth science, human space exploration, and study of the universe. This year's talks and demonstrations will reveal some of the innovative technologies behind projects such as the Curiosity rover's first year on Mars; designing cleaner, quieter aircraft; advancing global climate research and technologies; developing next-generation space vehicles and launch environments; understanding astrophysical phenomena; and approaches to data-intensive and quantum computing. Attendees can stop by the booth to experience spectacular scientific visualisations displayed on a state-of-the art 11-foot hyperwall.
The National Center for Supercomputing Applications (NCSA) (#1122), located at the University of Illinois at Urbana-Champaign, provides computing, data, networking, and visualisation resources and services that help scientists and engineers better understand our world. At SC13, NCSA will showcase its capabilities and projects in big data, extreme-scale computing, collaborative e-science, and industry collaboration.
NCSA is home to the NSF-supported Blue Waters petascale supercomputer, which is available to scientists, engineers, educators, and companies across the US for large-scale modelling and simulation. NCSA’s exhibit will include information about Blue Waters projects and successes as well as how to apply to use the system; Blue Waters fellowships are also available to graduate students. Other showcased projects include the Extreme Science and Engineering Discovery Environment (XSEDE), scientific visualisation, and a $10 million NSF-funded effort to create software to manage and make sense of vast amounts of uncurated digital data.
Numascale (#2505) is a leader in interconnect technology development with its low-cost scalable shared memory adapter for commodity servers. Numascale is the first 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 one single system running one instance of the operating system. The coherent cache reduces latency to remote memory and provides cache coherency on a cache line level. This caching scheme aligns perfectly with how the higher level caching in the system utilises standard optimisations in system software. Utilising directory-based caching, systems are scalable from two to thousands of servers connected using 2D or 3D torus topologies and up to 256TB of memory. At SC13, Numascale will feature an installation with 1,728 cores and 4.6 Tbytes of memory.
The Numerical Algorithms Group (NAG) (#1716) delivers trusted numerical computing software and high-performance computing services. NAG experts work closely with researchers in academia and industry to create powerful, accurate and flexible software which is relied upon by tens of thousands of users, companies, and learning institutions, as well as numerous independent software vendors.
From multicore workstations to the latest supercomputers, NAG's robust software development products, parallel software engineering services, independent technology consulting and HPC training can support modelling and computing needs. Visitors to NAG’s booth can find out about the latest HPC projects, the routines in the recent release of the NAG Library, NAG’s support for the latest co-processors, and NAG’s other numerical services.
At SC13, Nvidia (#613) is exhibiting what it says is game-changing scientific discovery and innovations made possible by GPU-accelerated computing. On display will be the newest GPU-accelerated technologies and examples of how scientists are advancing computational science.
Nvidia will also feature a GPU Technology Theatre with talks on a wide range of topics in high-performance computing and accelerated computing. This year's line-up includes speakers from Proctor & Gamble, ETH Zurich, Oak Ridge National Laboratory, GE Global Research, and many others.
Additionally, the company is presenting numerous tutorials and a developer zone will be available for anyone who is looking to get started with accelerating on GPUs. GPU-accelerated computing will also be on display from partners including ARM, Cineca, Nice, OpenACC, SGI, HP, Dell, IBM, Supermicro, and many others.
OpenCL (#4137) is a widely-supported open standard for programming heterogeneous computers built from CPUs, GPUs, FPGAs, and other processors. It includes a framework to define the platform as a host and one or more devices, plus a C-based language for writing programs that run on the devices. OpenCL programs can therefore use all the resources of the heterogeneous computer.
At SC13, the latest developments in OpenCL will be shown, including the new OpenCL 2.0 specification. This major upgrade to OpenCL adds support for next-generation heterogeneous systems, including shared virtual memory to share addresses between the host and the devices; support for nested parallelism enabling kernels to directly enqueue more kernels; a formal memory model based on C11, and much more. At the BOF session, discussions will revolve around how programmers can get more involved with OpenCL, including an OpenCL workshop (iwocl.org) planned for 12-13 May 2014 in Bristol UK.
The OpenMP Consortium (#4117) will present OpenMP 4.0 at SC13. OpenMP 4.0 is a major advance that adds two new forms of parallelism in the form of device constructs for the support of accelerators, and SIMD constructs to vectorise both serial as well as parallelised loops. It also includes several significant extensions for the loop-based and task-based forms of parallelism already supported in OpenMP 3.1.
There will be presentations at booth #4117, a BOF on OpenMP and an Exhibition Forum presentation focused on OpenMP 4.0. Jointly defined by a group of major computer hardware and software vendors and the user community, OpenMP is a portable, scalable model that gives shared-memory parallel programmers a simple and flexible interface for developing parallel applications for platforms ranging from multicore systems and SMPs, to embedded systems.
Attendees are invited to meet the PGI (#2509) compiler team at SC13. This year, PGI will be demonstrating the upcoming PGI 2014 release packed with many new features and performance enhancements including industry-leading OpenMP for multi-core x64 CPUs, full support for OpenACC on both Nvidia Tesla and AMD Radeon GPUs, support for Cuda Fortran debugging on Nvidia GPUs, and a new freely available (as in free!) version of its optimising PGI Fortran 2003 and C99 compilers for MacBook. Visitors to the booth will be able to watch the three demonstrations and receive a custom PGI T-shirt, with a limit of one per attendee, while supplies last.
PRACE (#3341), the Partnership for Advanced Computing in Europe, will be showcasing its globally competitive high-performance computing ecosystem for European scientific and industrial competitiveness. PRACE is a pan-European Research Infrastructure that spans 25 nations and pools resources, leading-edge supercomputers as well as HPC expertise. The 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 SC13 will be the PRACE Open R&D access for industry and the PRACE SHAPE pilot (SME HPC Adoption Programme, www.prace-project.eu/shape) for small- to medium-sized businesses, as well as the ambitious Dare to Think the Impossible (www.daretothinktheimpossible.com) outreach campaign to educate tomorrow's researchers about the power of supercomputing and how it is applicable to society’s grand challenges. Visitors will also be informed of the training offered by the PRACE Advance Training Centres (PATCs – www.training.prace-ri.eu).
Q-Leap Networks (#641) will present the newest version of Qlustar, its full-fledged HPC/storage cluster operating system (OS). Release highlights include a next-generation remote execution engine that permits easy and fast execution of parallel commands on cluster nodes, powerful real-time analysis of completion and exit status as well as command output grouping. In addition, a completely revised Lustre/ZFS stack includes the newest high-availability components capable of supporting the most demanding HPC or big data requirements.
Qlustar is designed to scale from small departmental clusters to multi-thousand node supercomputers. Its core is a lightweight, modular compute/storage-node OS, and as a whole, it is a complete Linux distribution based on Debian/Ubuntu. QluMan, Qlustar's management and monitoring framework, ensures easy and consistent cluster setup and operation, and keeps track of hard and software failures, and more. With Qlustar, node provisioning belongs to the past – simply assign an OS image to a node and (re)boot.
Demonstrations of Networking Infrastructure as a Service (NIaaS) and presentations on tools for managing large distributed data sets and keeping data safe will be the focus of the RENCI/North Carolina booth (#4305) at SC13. A demonstration of the experimental ExoGENI infrastructure, which allocates network bandwidth and cloud-based computing and storage resources on demand in order to run scientific workflows, will be featured in the booth and as part of the SCinet Network Research Exhibition.
Information about the integrated Rule-Oriented Data System (iRODS), including tutorials, and updates on the new iRODS Consortium, will also be presented. Other data solutions to be highlighted in the booth include The Secure Research Space, which provides secure, cloud-based spaces for working with sensitive data, and DataBridge, a project designed to make connections among scientific datasets in the same way Facebook makes connections between people and social data.
RSC Group (#404), a leading Russian innovative HPC solutions builder, will demonstrate its patented, energy-efficient and scalable RSC Tornado architecture-based HPC and data centre solutions at SC13. In addition to featuring direct liquid cooling on widely available commodity server boards (for example, Intel S2600JF) and top Intel Xeon E5-2697 v2 processors, Intel SSD and high-speed InfiniBand FDR interconnect, a current industry-record 211 Tflops per rack performance is delivered with Intel Xeon Phi 7120X coprocessors, PUE 1.06 and 100+ kW/rack power density.
RSC offers HPC/enterprise solutions from mini DC to exascale-ready Petaflop supercomputers with hundreds kW/rack, providing for 60 per cent power savings and an impressive 92 per cent computing efficiency ratio by Linpack in most custom flexible and green design. At SC13, RSC will also launch what it says is a revolutionary ultra-high density liquid-cooled HPC product that will deliver a new breakthrough with increased performance per rack.
Scalable Informatics (#1919) is a provider of tightly-coupled fast computing, storage, and networking appliances for time series, massive storage, data warehouse, and noSQL/SQL databases. The innovative design — tightly coupling storage and computing in the same unit — eliminates bottlenecks to deliver the highest performance at aggressive price points. Scalable will be demoing the si-DWA at SC13, a high performance scalable parallel SQL plug-and-play appliance for massive databases.
Running the Xtreme dbX ANSI SQL database, the si-DWA offers the highest performance per watt and per square foot in the industry, and is designed for high performance, ease of use, and quick deployment, delivering rapid query response times for today’s ever-growing online big data warehouse environment. The si-DWA bypasses the learning curve by supporting standard SQL, and is data schema/model agnostic. The system comes fully installed with all software and is ready for analysing user data.
ScaleMP (#3732) delivers a platform designed for environments that process complex or unstructured data, or data that requires the use of multiple algorithms. vSMP Foundation is a software-based SMP that holds data and enables different processes to operate on the same data while in memory. By retaining more of the frequently used data in memory, the overall analytics capability and speed to results is increased.
The vSMP Foundation product suite aggregates multiple, industry-standard, off-the-shelf x86 servers into one single virtual high-end system. Hundreds of customers use the vSMP Foundation suite of products as an alternative to traditional, proprietary, and expensive symmetrical multiprocessing (SMP) systems, benefiting from the cost structure of x86 clusters. vSMP Foundation also provides cluster users and administrators with the simplified clustering infrastructure of a single operating system. vSMP Foundation scales up to 256TB of memory and up to 32,768 CPUs.
Silicon Mechanics (#3126) is looking forward to participating in the 25th anniversary of SC13 to highlight its ongoing commitment to state-of-the-art computing in higher education and research. One way the company gives back to the education and research community is by providing a complete high-performance computer cluster as part of a highly competitive research grant programme. This year’s cluster, valued at more than $118,000, includes hardware and software from Intel, sTec, Mellanox, Nvidia, Kingston, Seagate, Supermicro, Bright Computing, and LSI Logic.
For the third year in a row, Silicon Mechanics is sponsoring students from the Massachusetts Green HPC Consortium (MGHPCC) team in SC13’s Student Cluster Competition. MGHPCC includes Boston University, Harvard University, MIT, University of Massachusetts, Tufts University, and Northeastern University. Silicon Mechanics identifies hardware partners, assembles and tests the cluster, and then ships it for use in the competition.
Spectra Logic (booth #832) will be showcasing its T-Finity and T950 Tape Libraries as well as its nTier Verde ArchiveGrade Disk Product and Black Pearl Deep Storage Appliance.
Spectra defines, designs and delivers innovative data protection through tape and disk-based backup, recovery and archive storage solutions. By igniting innovation it challenge expectations of the data protection market with intelligent, integrated, and simple to use backup and archive technologies. With more than 30 years of experience in storage, Spectra's goal is to develop high-density, feature-rich storage products with service and support to customers worldwide. To guarantee superior quality for clients, all T-Series libraries are manufactured onsite at headquarters based in Boulder, Colorado, USA.
Supermicro (#3132) takes its 2U Twin architecture to the next level of performance and expandability with the high efficiency 2-node TwinPro and high density 4-node TwinPro². TwinPro architecture provides the highest performance per-watt/per-dollar with energy efficient cooling-optimised designs, fans, heat sinks, redundant Platinum Level high efficiency power supplies and sophisticated power management software. Each node supports dual Intel Xeon E5-2600 v2 ‘Ivy Bridge’ processors, 16x DDR3 Reg. ECC DIMMs, a rear add-on-card expansion slot and 1x PCI-E 3.0 x16 and x8 LP slot (TwinPro) or 1x PCI-E 3.0 x16 slot (TwinPro²).
Each node features on-board dual 10GbE and QDR/FDR InfiniBand alongside a dedicated IPMI remote management port. TwinPro nodes support 8x SAS3 12Gb/s and 4x SATA hot-swap HDDs. TwinPro² nodes support 6x 2.5-inch SAS3 hot-swap HDDs. For ultimate performance, the 2-node TwinPro accommodates a double-width GPU and two add on cards per node.
SysFera (#2417) develops software that makes HPC simple to use. The team is happy to announce the latest release of SysFera-DS, the solution for HPC infrastructure management and remote application usage. This version boasts an improved remote-visualisation module, extended support for cloud computing, as well as statistics and reporting tools that allow billing to customers and business units.
SysFera is proud to announce that CINES (the French Computing Center for Higher Education) is now using the SysFera-DS solution to offer users improved access to their visualisation clusters. A customer feedback session presenting the positive impact of the solution will be hosted by Olivier Rouchon (CINES's head of the HPC department) and Marie Granier (SysFera's sales manager) during the SC13 Exhibitor's Forum on Wednesday, 20 November at 11.15am.
The Texas Advanced Computing Center (TACC) (#2732) designs, deploys, operates and conducts research and development to provide a comprehensive, integrated set of computational, visualisation, data analysis, storage, and software activities. Collectively, this set of balanced resources, interactive activities, and expert user support make TACC the Ultimate Science Environment (USE). Visit the booth to learn how TACC enables more open science across all disciplines than any other research centre in the US, and experiment with new visualisation technologies, including an Oculus Rift and Pico Wall. As always, TACC will offer t-shirts on a first-come, first-served basis to its guests. Ask how you can also receive a limited-edition laptop sticker ‘My other computer is super’ — the must-have accessory for SC13!
Univa (#4101), the data centre automation company, is a leading provider of automation and management software for computational and big data infrastructures. Univa is the sole commercial provider of Grid Engine software, and the company’s products and global enterprise support give its customers the power to manage their entire computational space, no matter how big or where it is deployed. Many of the leading brands in the world depend on Univa’s unsurpassed expertise, and premier services and support. Univa is headquartered in Hoffman Estates, Illinois, USA, with offices in Markham, ON, Canada; Austin, Texas, USA and Munich, Germany.
XSEDE (#422) is a virtual organisation that provides a dynamic distributed infrastructure, support services, and technical expertise that enable researchers, engineers, and scholars to address the most important and challenging problems facing the nation and world. Stop by the XSEDE booth at SC13 in Denver to find out how you can benefit from its computing systems, training, and other services to accelerate your research.
The XSEDE booth will allow visitors to interact with senior management personnel, including project director John Towns and key leadership from operations, user services, outreach, and more. Additionally, if visitors have a useful technology that needs more exposure, it can be entered immediately into the Technology Investigation Services’ XSEDE Technology Evaluation Database (XTED) for free to show the world what has been created.
Xyratex (#2107) is the number one OEM provider of enterprise data-storage solutions, including the award-winning ClusterStor family of high performance scale-out storage solutions. ClusterStor’s software architecture tightly integrates the Lustre file system with Xyratex’s enterprise-quality storage. The ClusterStor family delivers availability, ease of use, price-performance and overall total cost of ownership. For applications where users need storage that scales both in capacity and performance, ClusterStor supports high performance data processing from relatively small data quantities to some of the largest supercomputers in the world.