HPCwire's 2007 Readers' and Editors' Choice Awards
2007 Awards Presentation
2007 has been a watershed year for a number of companies in the HPC realm. New products and technologies were introduced that signal a change in the way high performance computing will be accomplished, while enabling new classes of applications.
The traditional boundaries that once limited computationally-intensive computing to the supercomputing market no longer exist. The rise of multicore processors and the continued dominance of cluster computing are spreading the reach of HPC farther each day. And, as the ecosystem evolves, new organizations are coming to the fore that demonstrates the vitality of a growing community.
We believe the 2007 HPCwire Readers' and Editors' Choice Awards, determined by a poll of our global readers, contributing editors, and industry luminaries, depicts the new face of the HPC community. The range of end-user organizations and vendors reflects the extent to which this community has become both mainstream and exceptional.
We are proud to present the collective opinion of the HPC community and we congratulate the winners of the 2007 HPCwire Readers' and Editors' Choice Awards.
Readers' Choice Award Recipient:
Boeing
Boeing is one of the world's leading aerospace companies and the largest manufacturer of commercial jetliners and military aircraft combined, with capabilities in rotorcraft, electronic and defense systems, missiles, satellites, launch vehicles and advanced information and communication systems.
Boeing could not be competitive in the commercial airplane market without extensive use of HPC. Even small, incremental performance improvements in our products have the potential to make a huge difference in the company's bottom line. Less design time using fewer resources (physical testing is very expensive in the aerospace world) is the end result of intelligent use of HPC resources. www.boeing.com
Editors' Choice Award Recipient:
Pratt & Whitney
Pratt & Whitney is one of the foremost designers and manufacturers of turbine engines for commercial and military applications. Their engines are installed in more than half of the world's commercial aircraft fleet. The company's military engines power the Air Force's F-15 and F-16s, and the F119 and F135 engines will power the front-line fighters of the future, the F-22 Raptor and Joint Strike Fighter.
High performance computing is used to drive Pratt & Whitney's mission to provide engines at the lowest possible cost and the highest level of reliability. From aerodynamic design to the simulation of bird impact to the design of "greener" engines, high performance computing has become an indispensable facet of the company's research and development efforts. www.pw.utc.com
Readers' Choice Award Recipient:
Ferrari
Ferrari is a symbol of technical excellence thanks to its reputation as a pioneer in automotive engineering. Car manufacture is carried out in the heart of Maranello and the all-aluminium chassis and bodies are built by Ferrari's specialist aluminium production arm, Carrozzeria Scaglietti, in nearby Modena. Ferrari currently builds over 4,500 cars a year with production split between the V8-engined cars, the F430 and F430 Spider, and the V12-engined cars, the 599 GTB Fiorano 2-seater berlinetta and the flagship 2+2 612 Scaglietti.
Ferrari constantly pushes the envelope of performance on its road cars thanks to its advanced aerodynamics studies, and this competitive advantage is today ensured by no less than three separate in-house wind tunnels. The most recent of these is the futuristic Formula 1 wind tunnel, designed by Renzo Piano, which opened in 1997.
CFD simulation empowers Ferrari teams to develop the concept of their car in a complementary way with their wind tunnel facilities in the year preceding each season. During the racing year, they use these tools to improve the car for each race circuit. www.ferrariworld.com
Editors' Choice Award Recipient:
BMW
With the three brands, BMW, MINI and Rolls-Royce Motor Cars, the BMW Group has its sights set firmly on the premium sector of the international automobile market. To achieve its aims, the company knows how to deploy its strengths with an efficiency that is unmatched in the automotive industry. From research and development to sales and marketing, BMW Group is committed to the very highest in quality for all its products and services. The company's phenomenal success is proof of this strategy's correctness.
The BMW Sauber F1 Team employs state-of-the art supercomputing resources to optimize CFD aerodynamics for its Formula 1 automobiles. Aerodynamics has a crucial influence on the performance of modern Formula 1 cars, with experimental work in the wind tunnel and computational fluid dynamics complementing each other. www.bmw.com
Readers' Choice Award Recipient:
National Resource for Biomedical Supercomputing (NRBSC) / Pittsburgh Supercomputing Center (PSC)
The National Resource for Biomedical Supercomputing (formerly the Biomedical Initiative) was established at the Pittsburgh Supercomputing Center in 1987 as the first extramural biomedical supercomputing program in the country funded by the National Institutes of Health. NRBSC pursues leading edge research in high performance computing and the life sciences, and fosters exchange between PSC expertise in computational science and biomedical researchers nationwide. The organization provides computational biomedical research and outreach to the national biomedical research community through education and publications. Research at NRBSC is centered in three areas: microphysiology; volumetric visualization and analysis; and computational structural biology. www.nrbsc.org
Editors' Choice Award Recipient:
Boston Scientific
For more than 25 years, Boston Scientific has advanced the practice of less-invasive medicine by providing a broad and deep portfolio of innovative products, technologies and services across a wide range of medical specialties. These less-invasive medical technologies provide alternatives to major surgery and other medical procedures that are typically traumatic to the body. In less-invasive procedures, devices are usually inserted into the body through natural openings or small incisions and can be guided to most areas of the anatomy to diagnose and treat a wide range of medical problems. As a pioneer in medical devices delivering more than 15,000 products to clinicians in over 45 countries, Boston Scientific is building the knowledge, processes and infrastructure it needs to identify opportunities, address new disease areas and support clinicians in advancing the practice of health care. www.bostonscientific.com
Readers' Choice Award Recipient:
BNP Paribas
BNP Paribas is a European leader in global banking and financial services and is one of the 5 strongest banks in the world according to Standard & Poor's. The group is present in over 85 countries, with 155,000 employees, including 123,000 in Europe. The group holds key positions in three major segments: corporate and investment banking, asset management & services and retail banking. Present throughout Europe in all of its business lines, the bank's two domestic markets in retail banking are France and Italy. BNP Paribas also has a significant presence in the United States and strong positions in Asia and the emerging markets. www.bnpparibas.com
Editors' Choice Award Recipient:
III Offshore Advisors
Hedge fund manager III Offshore Advisors is using Digipede's distributed computing solution on top of Microsoft's .NET environment to measure risk in credit derivative portfolios. The hedge fund manager trades non-dollar instruments in Japan and Europe and employs the grid to run "what if" scenarios. The quick turnaround time afforded by it high performance computing platform allows the team at III Offshore Advisors to concentrate on identifying and executing the best trades faster than its competition.
Readers' Choice Award Recipient:
Petroleum Geo-Services
Petroleum Geo-Services (PGS) is a leading geophysical company, which provides an extensive range of seismic services and products for the petroleum industry including data acquisition, processing, reservoir analysis and interpretation. PGS operates 12 marine streamer vessels including 6 vessels of the unique Ramform class. The company also operates between 7 and 10 Onshore crews and has 15 data processing centers.
Since the start of the company in 1991, PGS has pioneered the development of multi-streamer marine seismic acquisition, producing increasingly efficient, high-quality 3D seismic data for the industry. The company has also introduced high-density 3D seismic (HD3D) in all environments and developed in-house expertise in geology, geophysics, and reservoir analysis. PGS also provides onshore seismic services where the company has a reputation for using the latest equipment in challenging environments. The data processing capabilities of PGS have grown substantially from originally processing PGS data to a major player in a high technology industry. www.pgs.com
Editors' Choice Award Recipient:
Shell Oil Company
Shell's technology is geared to produce energy and petrochemicals sustainably and economically. In the face of 21st century demand, the company is applying the latest technological solutions for exploring and producing oil and gas, processing and refining products, and developing energy from new sources.
Shell is applying high performance computing technology and advanced visualization to help discover new oil fields and get the most of existing ones. The company's 12 virtual reality centers help engineers and scientists collaborate in real time to help study and evaluate reservoir potential. With global demand for oil rapidly increasing and easy oil becoming scarcer, Shell is beginning to explore more difficult reservoirs from which to produce our future hydrocarbons. www.shell.com
Readers' Choice Award Recipient:
Sony Pictures Imageworks
Sony Pictures Imageworks state-of-the-art visual effects and character animation company dedicated to the artistry of digital production and character creation. The company has been recognized by the Academy of Motion Picture Arts and Sciences with Oscars for Achievement in Visual Effects for Spider-Man 2 and Best Animated Short Film for The ChubbChubbs! as well as nominations for Superman Returns, Monster House, The Chronicles of Narnia: The Lion, the Witch and the Wardrobe, Spider-Man, Hollow Man, Stuart Little and Starship Troopers. Imageworks continually raises the bar in the visual effects and character animation industry by providing leading-edge technology to world-class artists. www.imageworks.com
Editors' Choice Award Recipient:
DreamWorks Animation
DreamWorks Animation is devoted to producing high-quality family entertainment through the use of computer-generated (CG) animation. With world-class creative talent and technological capabilities, the company's goal is to release two CG animated feature films a year that deliver great stories, breathtaking visual imagery and a sensibility that appeals to both children and adults. In 2004, DreamWorks Animation SKG became the first animation company to produce and distribute two CG animated features in a single year, including Shrek 2, the third highest-grossing movie of all time. www.dreamworksanimation.com
Readers' Choice Award Recipient:
LANL / IBM / AMD for Roadrunner
The "Roadrunner" supercomputer at the U.S. Department of Energy's Los Alamos National Laboratory represents a first-of-a-kind design, Cell B.E. chips - originally designed for video game platforms - will work in conjunction with systems based on x86 processors from AMD. Designed specifically to handle a broad spectrum of scientific and commercial applications, the supercomputer will include highly sophisticated software to orchestrate over thousands Opteron and Cell processors in tackling some of the most challenging problems in computing today. The revolutionary supercomputer will be capable of a peak performance of over 1.6 petaflops. www.lanl.gov/roadrunner/
Editors' Choice Award Recipient:
DOE INCITE Program
Over the past 30 years, the Department of Energy's (DOE) supercomputing program has played an increasingly important role in scientific research by allowing scientists to create more accurate models of complex processes, simulate problems once thought to be impossible, and to analyze the increasing amount of data generated by experiments. To help the research communities fully tap into the capabilities of current and future supercomputers, Under Secretary for Science Raymond Orbach launched the Innovative and Novel Computational Impact on Theory and Experiment (INCITE) program in 2003. The INCITE program was conceived specifically to seek out computationally intensive, large-scale research projects with the potential to significantly advance key areas in science and engineering. hpc.science.doe.gov
Readers' Choice Award Recipient:
Rackable Systems
Rackable Systems offers server and storage products for large-scale data center deployments, focusing on computational density, efficiency, performance, thermals and lights-out management to help customers achieve lowest TCO. The company's built-to-order servers and storage in x86 1U, 2U, 3U and 4U form factors are configured in industry-first, half-depth, "back-to-back" rack mounting. In addition to improved system density, the platforms feature reduced power consumption, better cooling designs and simplified serviceability. www.rackable.com
Editors' Choice Award Recipient:
AMD Quad-Core Opteron
AMD's quad-core Opteron processors with Direct Connect Architecture are designed for superior levels of multi-threaded performance, power efficiency, and low TCO - all within a consistent footprint, and power and thermal envelopes. The native quad-core design features an enhanced cache structure and integrated memory controller designed to sustain high application throughput. With its outstanding memory bandwidth, system scalability, and performance-per-watt, the new AMD quad-core Opteron has become a favorite for many high performance computing applications. www.amd.com
Readers' Choice Award Recipient:
Panasas ActiveStor
The Panasas ActiveStor Storage Clusters are designed to meet the demands of a massively parallel data movement by offering scalable performance and capacity via an object-based approach to Storage. For environments that integrate batch and interactive applications the ActiveStor 5000 series provides these capabilities in a unified storage solution, eliminating the need for multiple storage systems. Where batch processing is the driving consideration the ActiveStor 3000 series delivers high capacity bandwidth and throughput, enabling jobs to be processed as quickly as possible. With simple management and a single virtual namespace, Panasas Storage Clusters accelerate time to results for engineers and scientists in both batch and interactive environments. www.panasas.com
Editors' Choice Award Recipient:
Sun Fire x4500 Server
The Sun Fire X4500 Server, code-named Thumper, brings together state-of-the-art server and storage technologies in a single box to deliver high-performance I/O. By integrating server and storage capabilities, the Sun Fire X4500 Server delivers the high performance of a four-way x64 server and exceptional storage density, with up to 48 TB in 4U of rack space. This system also delivers high data throughput for about half the cost of traditional solutions. www.sun.com
Readers' Choice Award Recipient:
HP Scalable Visualization Array
The HP Scalable Visualization Array (HP SVA) is a scalable, ready-to-run visualization solution that completes the HP Unified Cluster Portfolio's integration of computation, data management and visualization in a single, integrated cluster environment. The HP SVA solution adds high performance HP Workstations in building block configurations that combine with industry-standard visualization components and integrate with the HP XC System Software and HP StorageWorks Scalable File Share.
The tight integration of scalable computation, data management and visualization enables clustered parallel visualization applications with support for very large data sets; display of complex, high resolution images, including volume visualization; and real-time rendering, with computational steering via closely coupled visualization, computation and data management. www.hp.com
Editors' Choice Award Recipient:
HP's Parallel Compositing Library
The HP Parallel Compositing Library provides a set of software functions that facilitate development of parallel graphics applications. The library provides mechanisms for describing the desired compositing, the participating nodes contributing partial images, input data, and desired output. It handles communication and synchronization between nodes, and implements logic for combining pixels from partial images - produced by individual graphic nodes - to properly construct a complete image for each frame. The library also provides extensibility for creating new compositing operators.
The HP Parallel Compositing Library benefits scientists, researchers and engineers doing complex and demanding scientific and industrial visualization in segments such as scientific research, CAE, O&G, defense, life-sciences, medical-imaging and DCC/animation. www.hp.com
Readers' Choice Award Recipient:
Mercury Multicore Plus SDK for PS3
The MultiCore Plus Software Development Kit (SDK) is a suite of software products specifically designed for next-generation multicore processors such as the Cell Broadband Engine (BE) processor. The SDK includes a comprehensive programming framework, highly optimized math libraries, and a powerful debug and analysis tool.
The focus of the MultiCore Plus SDK is application performance and developer productivity. This seamless package of software development tools and libraries helps users take full advantage of the Cell BE processor's architecture to maximize application resources and boost performance. www.mc.com
Editors' Choice Award Recipient:
Interactive Supercomputing's Star-P
The Star-P software platform enables scientists, engineers and analysts to transparently use high performance computing resources, using familiar desktop tools. Star-P software is a client-server parallel computing platform that's been designed to work with multiple high level client applications such as MATLAB, Python, or R, and has built-in tools to expand computing capability through addition of libraries and hardware-based accelerators. Star-P is part of a new generation of HPC software development platforms that is expanding end-user accessibility by targeting developer productivity. www.interactivesupercomputing.com
Readers' Choice Award Recipient:
Voltaire
Voltaire provides networked computing solutions for the next generation data center in the enterprise as well as government and research institutions. The company focuses on improving data center efficiency by enabling grid and cluster computing deployments on commodity server and storage offerings. Voltaire's family of switching hardware and network virtualization software is designed to deliver the highest performing, intelligent backbone for grid computing architectures. Using the InfiniBand standard, Voltaire solutions offer improved performance, utilization and scalability across compute clusters, storage and IP networks. www.voltaire.com
Editors' Choice Award Recipient:
Mellanox ConnectX
Mellanox Technologies supplies semiconductor-based, high-performance interconnect products that facilitate data transmission between servers and storage systems. Their products are an integral part of computing, storage and communication applications used in enterprise data centers, high-performance computing and embedded systems.
Mellanox ConnectX architecture, implemented on a single adapter chip, offers the building blocks for designing server and storage connectivity solutions for either InfiniBand or Ethernet applications. The ability to support multiple network and storage connectivity options on a ConnectX adapter enhances the server OEMs target markets and time to market. Finally, end users can use ConnectX-based solutions for incremental and demand-based I/O growth and scaling. www.mellanox.com
Readers' Choice Award Recipient:
Cisco
Cisco Systems Inc. is a worldwide leader in networking for the Internet. Today, networks are an essential part of business, education, government and home communications, and Cisco's Internet Protocol-based (IP) networking solutions are the foundation of these networks. Cisco was founded in 1984 by a small group of computer scientists from Stanford University. Since the company's inception, Cisco engineers have been leaders in the development of Internet Protocol (IP)-based networking technologies. Today, with more than 47,000 employees worldwide, this tradition of innovation continues with industry-leading products and solutions in the company's core development areas of routing and switching, as well as in advanced technologies such as IP communications, network security, wireless LAN, storage area networking, home networking, video systems, and application networking services. www.cisco.com
Editors' Choice Award Recipient:
Woven Systems Ethernet Fabric Switch
Woven Systems is a network infrastructure provider focused on designing and building massively scalable Ethernet fabric switching solutions for enterprise and Internet data centers. The company's Ethernet Fabric Switch technology uses dynamic congestion management to automatically balance traffic, delivering levels of performance on par with InfiniBand technology. Woven's switches incorporate patented technology, implemented in Woven's vSCALE packet processing ASIC, that enables non-blocking multi-path interconnectivity using standard Ethernet protocols. www.wovensystems.com
Readers' Choice Award Recipient:
Voltaire
Voltaire provides networked computing solutions for the next generation data center in the enterprise as well as government and research institutions. The company focuses on improving data center efficiency by enabling grid and cluster computing deployments on commodity server and storage offerings. Voltaire's family of switching hardware and network virtualization software is designed to deliver the highest performing, intelligent backbone for grid computing architectures. Using the InfiniBand standard, Voltaire solutions offer improved performance, utilization and scalability across compute clusters, storage and IP networks. www.voltaire.com
Editors' Choice Award Recipient:
Dell PowerEdge Servers
Dell high performance computing clusters help solve some of the most challenging computational intensive tasks facing business, educational and scientific communities. By integrating the latest advances in industry-standard servers, high speed interconnects and leading open source and commercial software, Dell's HPC clusters deliver the performance of proprietary SMP (symmetric multiprocessing) systems, with the simplicity and value of industry standard computing. www.dell.com/hpcc
Readers' Choice Award Recipient:
IBM Blue Gene/P
The IBM Blue Gene/P system is the next generation to the successful Blue Gene/L, and adheres to the original key design points set forth in the Blue Gene/L solution. Specifically, the Blue Gene family of supercomputers has been designed to deliver ultrascale performance within a standard programming environment while delivering efficiencies in power, cooling and floor-space consumption. Blue Gene/P extends the performance through increased computational density, 4-way SMP enhanced functionality, scalability for petaflop performance, and aggressive power management for low power consumption first established with the Blue Gene Solution. www-03.ibm.com/servers/deepcomputing/bluegene.html
Editors' Choice Award Recipient:
NVIDIA Tesla
NVIDIA Tesla is the industry's first GPU product line specifically targeted to high performance computing applications. A dedicated, high performance computing solution, Tesla brings supercomputing power to any workstation or server and to standard, CPU-based server clusters. Paired with CUDA, a C-language development environment for the GPU, Tesla offers scientists, engineers and other technical professionals access to the power of GPU computing. www.nvidia.com/object/tesla_computing_solutions.html
Readers' Choice Award Recipient:
PSC ZEST
ZEST, developed by Pittsburgh Supercomputing Center's Advanced Systems Group, is a prototype distributed file-system infrastructure that vastly accelerates write bandwidth on large compute platforms, alleviating a common bottleneck. ZEST will directly apply to petascale platforms and is designed to allow maximum reliable performance per unit cost for these systems. www.psc.edu
Editors' Choice Award Recipient:
Panasas Tiered Parity
As disk drive density increases, new challenges in reliability and data integrity have come to the fore. Panasas Tiered Parity is an innovative storage architecture designed to addresses these challenges via a three-tiered system of of error detection and correction. Each tier operates independently of the others allowing the error detection and correction codes to be individually tailored for that tier. The three tiers are complimentary to each other and collectively provide a comprehensive and scalable architecture for high performance storage. www.panasas.com/tiered_parity.html
Readers' Choice Award Recipient:
NVIDIA CUDA
NVIDIA CUDA technology is a new computing architecture that enables the GPU to solve complex computational problems in consumer, business, and technical applications. CUDA (Compute Unified Device Architecture) technology gives computationally intensive applications access to the tremendous processing power of NVIDIA graphics processing units (GPUs) through a new programming interface. Providing orders of magnitude more performance and simplifying software development by using the standard C language, CUDA technology enables developers to create innovative solutions for data-intensive problems. For advanced research and language development, CUDA includes a low-level assembly language layer and driver interface. www.nvidia.com/cuda
Editors' Choice Award Recipient:
RapidMind Development Platform
RapidMind is a framework for expressing data-parallel computations from within C++ and executing them on multicore processors. Developers of HPC and enterprise software are using RapidMind's platform to create single-threaded applications that leverage the potential of multicore processors from AMD and Intel as well as the acceleration capabilities available from GPUs and the Cell Broadband Engine. The framework allows developers to write code in standard C++, letting the RapidMind platform parallelize the application on the underlying hardware. www.rapidmind.net
Readers' Choice Award Recipient:
CEI EnSight
CEI offers a complete suite of engineering and scientific visualization tools, from meshing to plotting to animation, on all major operating systems. The company's products can be run on everything from laptops to workstations, clusters and supercomputers, with animations displayed in stereo and in immersive VR. The Ensight product line encompasses one of the largest feature sets of any visualization/post-processing tool on the market today. www.ensight.com
Editors' Choice Award Recipient:
Mercury Computer Systems OpenRTRT
OpenRTRT real-time ray tracing software is a 3D image-rendering application that combines the high quality and physically correct results of ray tracing with the interactivity of conventional shaded surface 3D graphics. Ray tracing is based on complex mathematical algorithms and are used to produce images of unprecedented realism. OpenRTRT scales in quality and performance when the computation is distributed across additional PCs in a cluster or across the network. www.mc.com
Readers' Choice Award Recipient:
SGI Altix
The Altix ICE was designed from the ground up with high performance computing in mind. The integrated blade architecture was designed to minimize system overhead, and drive the best possible performance density for optimal use of data center resources. The architecture uses dual- and quad-core Intel Xeon Processor architecture with a unique board design that yields high computational performance density, delivering up the 512 processor cores in a single rack, and scaling to thousands of nodes to address the most challenging compute problems. www.sgi.com
Editors' Choice Award Recipient:
SiCortex
With a team of 30 people, SiCortex has created family of cluster computer systems that are optimized from the silicon up for HPC applications. The company avoided the cost burden of a full custom design by leveraging the ecosystem of silicon manufacturers and suppliers of silicon intellectual property that has grown up over the past decade to serve vendors of specialized products such as cell phones and graphics processors. The result is a computer architecture that delivers high sustained performance in an extremely energy-efficient package. www.sicortex.com
Readers' Choice Award Recipient:
TeraGrid
Using high-end network infrastructure, the TeraGrid integrates high performance computers, data resources and tools, and high-end experimental facilities around the country. Currently, TeraGrid resources include more than 250 teraflops of computing capability and more than 30 petabytes of online and archival data storage, with rapid access and retrieval over high-performance networks. Through its Education, Outreach and Training (EOT) program, the TeraGrid is pro-actively engaging large numbers of, and more diverse communities of, researchers, educators and students to learn about, use, and contribute to applications of cyberinfrastructure in order to ensure America's long-term global competitiveness. www.teragrid.org
Editors' Choice Award Recipient:
U.N. Intergovernmental Panel on Climate Change
The IPCC was established to provide the decision-makers and others with an objective source of information about climate change. Based on results gleaned from sophisticated climate modeling work, the organization has helped shaped world opinion on the challenges of global warming. In 2007, the IPCC and former vice-president Al Gore were awarded the Nobel Peace Prize for their efforts in raising awareness of man-made climate change and in laying the foundation for possible solutions. www.ipcc.ch
Readers' Choice Award Recipient:
Dr. Eng Lim Goh (SGI)
Dr. Eng Lim Goh has been with SGI for 20 years, becoming one of the chief scientists in 1998 and chief technology officer in 2001. His tenure includes work in computer graphics algorithms and high performance computing (HPC) architectures. In HPC, he oversees Project Ultraviolet, the goal of which is to design and build the company's next generation science-driven computer architecture. He is also the coauthor of SGI's recommendation to the high-end computing revitalization task force (HECRTF) for federal funding of key corresponding technologies. This proposal was reviewed by HECRTF in 2003 and judged to be one of the top submitted papers. Dr. Goh is known as a proponent of next-generation computer systems designed specifically for customer applications performance. To this, he advocates computational density and a balanced multi-paradigm approach, across a globally addressable memory, to architectural design.
Editors' Choice Award Recipient:
Dan Reed (Microsoft Corp.)
As one of the most respected voices in the high performance computing community, Daniel Reed has been a vocal proponent of the importance of science and technology for American competitiveness. As chair of the President's Information Technology Committee (PITAC), he co-authored the 2004 report "Computational Science: Ensuring America's Competitiveness," which highlighted the strategic importance of computational science in every sector of the economy. As a member of the President's Council of Advisors on Science and Technology (PCAST), he helped draft an assessment of the NITRD program: "Leadership Under Challenge:Information Technology R&D," which outlines the actions needed to keep the U.S. in the leadership position in information technology. After a four-year tenure as director of the Renaissance Computing Institute Reed is now with Microsoft Corp. to head the company's scalable and multicore computing research initiative.
Readers' Choice Award Recipient:
IBM
IBM collaborates with clients to apply powerful, innovative deep computing solutions to some of today's most challenging and complex problems - enabling businesses and researchers to get results faster and gain a sustainable business advantage and to push the boundaries of scientific discovery. IBM remains the leader on the TOP500 list of supercomputers due to its ability to offer clients the broadest portfolio of deep computing solutions today. For more information visit www.ibm.com/deepcomputing.
Editors' Choice Award Recipient:
IBM
IBM collaborates with clients to apply powerful, innovative deep computing solutions to some of today's most challenging and complex problems - enabling businesses and researchers to get results faster and gain a sustainable business advantage and to push the boundaries of scientific discovery. IBM remains the leader on the TOP500 list of supercomputers due to its ability to offer clients the broadest portfolio of deep computing solutions today. For more information visit www.ibm.com/deepcomputing.
Having put some comfortable distance between itself and arch-rival AMD, Intel appears poised to dominate the x86 space for 2008 and beyond. The company's latest dual- and quad-core Xeon processors on their industry-leading 45nm process technology are already proving to be popular for high performance computing platforms. The Nehalem processor family, scheduled to be introduced in the second half of the 2008, should close the gap with AMD's more scalable architecture and give HPC system vendors a new reason to embrace Intel.
But the Intel has been here before and needs to avoid becoming complacent. The near-term challenges include getting the Nehalem processors to market on schedule, keeping its Larrabee effort on track in order to have a GPU computing solution to go up against AMD and NVIDIA, and shifting its adapting it product mix to meet the rising demands for mobile and high-end servers. Longer term, Intel is looking to coalesce its terascale computing strategy for the next generation of applications targeted for manycore processors. Meanwhile AMD hopes to outflank its larger rival once again, this time with integrated CPU-GPU processors and more power-efficient designs. www.intel.com
With this year's departure of founder Bill Gates from the day-to-day running of the company, Microsoft will symbolically cast off the PC-centric focus that dominated product strategy there for the first 30 years. This transition in leadership is taking place against a backdrop of a two paradigm shifts: from single-core computing to multicore computing, and from PC-based computing to utility computing. Microsoft is heavily committed to meeting both of those challenges, but it remains to be seen how the company will reshape itself inside this new reality.
Against this backdrop, Microsoft is in the midst of its pursuit of the high performance computing market. With the introduction Windows Compute Cluster Server (CCS) 2003 in 2005, the company offered the first Windows-compatible OS platform for the HPC market. This year they will introduce the sequel: Windows Server HPC 2008. Designed as a integrated OS/workload manager for smaller clusters, Microsoft is targeting their solution at the volume end of the market, where much of the HPC growth is being forecast. Here the challenge for Microsoft will be to prove to the open source-loving culture of the HPC community that there is still a place for proprietary system software in the ecosystem. www.mircosoft.com
Taking advantage of AMD's disarray in the acquisition of ATI and Intel's delayed entry into the high-end GPU processor market, NVIDIA was able to establish itself as a early leader in GPU computing in 2007. With its CUDA programming environment and Tesla computing products, the company introduced the first solutions specifically targeted for HPC applications. In selected technical computing applications, GPUs are able to achieve one or two orders of magnitude of performance improvement over CPUs.
By leveraging the intellectual property from a high-volume processor business and by providing a device-independent programming interface in CUDA, NVIDIA is able to demonstrate some compelling advantages over other HPC accelerator solutions. Its challenge will be to maintain technological leadership against rivals AMD and Intel and continue to evolve its own product line for HPC market. www.nvidia.com
The RapidMind platform aims to take the pain out of parallel computing on mainstream multicore and accelerator processors by offering a high-level software development environment. With the virtual demise of PeakStream at the hands of Google, the RapidMind platform became the only software solution that wrapped x86 multicore, GPU computing, and Cell processor acceleration into a single environment.
While many software solutions aimed at extracting concurrency from multicore processors and accelerator co-processors involve a tradeoff between ease of programming and performance, the RapidMind platform can achieve both. By leveraging existing C++ tools and using on an API that hides the complexities of hardware optimizations and load balancing underneath the hood, the platform is able to make best use of the underlying hardware without passing the programming burden onto the developer. Over the next two years, look for RapidMind to make its case through customer success stories. www.rapidmind.net
Under the guidance of CEO Jonathan Schwartz, CTO Greg Papadopoulos, and chief architect Andy Bechtolsheim, Sun Microsystems might finally have found the right balance of business sense, vision and innovation required for a viable HPC strategy. Sun's excitement about for utility computing, its embrace of open source software, its recently renewed relationship with Intel, demonstrates a willingness to be both a leader and follower when appropriate. The acquisition of the Lustre file system also shows the company has a sense of the strategic value of HPC technology in the broader enterprise market.
With its x86 Sun Fire rack mount servers and Sun blades, the company offers some solid HPC hardware. At the high end, Sun's success with their new petascale Constellation system will determine if the company has found a supercomputing architecture with legs. The performance of the new Ranger installation at the Texas Advanced Computing Center should give us some indication of architecture's viability. In the broader HPC business, Sun needs to be wary of its religious devotion to Solaris and its tendency to over-hype what are essentially well-engineered platforms. www.sun.com