“Industry leaders are turning to scientists for help in predicting how climate might affect their business or community in the coming decades. The University of Colorado at Boulder’s latest Dell supercomputer with high-performance Mellanox InfiniBand connectivity provides researchers and scientists with a highly efficient and scalable high-performance server and storage cluster with zero compromise. The new systems will help us to achieve high-resolution climate modeling and to improve accuracy in depicting regional atmospheric disturbances, and the opportunity to make accelerated progress on the world’s most important scientific problems.”

-Henry Tufo, Associate Professor of Computer Science and Computational Science Center Director at The University of Colorado at Boulder

“Our mission is to develop technologies that break down barriers for scientists working to solve some of society’s biggest challenges. Joined by Mellanox, we’re proud to be doing just that – making high-performance computing more efficient for Dr. Tufo and his team at the University of Colorado.”

-John Mullen, VP and General Manager for higher education at Dell

“There is a growing class of enterprise and cloud applications that have demanding time constraints and scaling requirements that are unable to be satisfied with general-purpose operating systems and I/O technologies. By delivering an integrated high-performing, standards-based Red Hat and Mellanox solution, we are enabling IT organizations to reduce costs, while providing the value of realtime access to information in broad markets, such as financial services, telecommunications and a large variety of real-time Java applications.”

-Tim Burke, vice president, Platform Engineering at Red Hat

“We are very proud to lead China’s effort in HPC and Cloud Computing technologies in catching up with the world. We are also pleased with the joint effort with our partner Mellanox for successfully building up the Nebula system. We certainly look forward to working together continuously in China’s market, to deliver outstanding performance, unprecedented reliability and superb efficiency to our mutual customers.”

-Jun Li, CEO and President of Dawning

“The combination of Mellanox 40Gb/s InfiniBand interconnects and GPU computing opens up a new world of possibilities to accelerate science and engineering research. Products coming to market employing the technologies of both NVIDIA and Mellanox help set the stage for next-generation, GPU-based clusters that are both high performance and highly efficient.”

-Andy Keane, general manager, Tesla business at NVIDIA

“As the popularity of GPU-based computing continues to increase, the importance of NVIDIA GPUDirect together with Mellanox’s offloading-based InfiniBand technology is critical to our world-leading HPC systems. We have implemented NVIDIA GPUDirect technology with Mellanox ConnectX-2 InfiniBand adapters and Tesla GPUs and have seen the immediate performance advantages that it brings to our high-performance applications. Mellanox offloading technology is an essential component in this overall solution as it brings out the real capability to avoid the CPU for the GPU-to-CPU communications. ”

-Dr. HUO Zhigang, The National Research Center for Intelligent Computing Systems (NCIC)

“By incorporating Mellanox 40Gb/s InfiniBand with NVIDIA GPUDirect technology, we have been able to conduct scientific simulations using GPUs at performance levels that we would never have been able to achieve using a different interconnect. The new Mole-8.5 Petaflop cluster, with industry-leading interconnect performance and efficiency, enables us to shorten the time it takes to run applications that are critical in the process of scientific discovery.”

-Dr. Xiaowei Wang of IPE

“IBM is delivering solutions that maximize application performance but also meet the requirements for infrastructure economics and green initiatives. Mellanox adapters and switches complement our Intelligent Cluster and iDataPlex products, giving IBM customers both performance and efficiency benefits that are critical ingredients for increased flexibility and higher return on their compute investment.”

-Roland Hagan, VP System X Servers at IBM

“The Best of Interop Awards highlight the outstanding IT vendors of the industry and this year’s Best of Interop is particularly exciting because of the strong competition we saw in each category. The selected winners represent innovation in the marketplace and deserve recognition for helping build the energy and growth of today’s IT Industry.”

-Art Wittmann, Managing Director of InformationWeek Analytics

“While FCoE will be an important I/O virtualization technology, customers have few choices in top-of-rack switches that can connect to existing FC SANs. Solutions that enable more vendors to provide FCoE-to-FC bridging in top-of-rack switches will increase customer choices, increase competition and help adoption of FCoE.”

-Joe Skorupa, Research VP, Gartner, Inc.

“Fast processing of exponentially growing market data feeds, split-second completion of analytics and low-latency trade executions are cornerstones of competitiveness for financial services institutions. NYSE Technologies Data Fabric messaging platform, which is used widely in this industry, can benefit from Mellanox’s ConnectX-2 with RoE solution to deliver extremely low latencies on Ethernet while scaling to handle millions of messages per second.”

-Feargal O’Sullivan, Head of Trading Solutions, Americas, at NYSE Technologies

“Cisco’s innovative Data Center 3.0 portfolio, including the Unified Computing System and Nexus switch technologies, was the first in the industry to deliver a unified fabric over lossless Ethernet to reduce IT infrastructure cost and complexity. Solutions that provide low-latency, deterministic performance and reliability such as the Mellanox ConnectX-2 EN with RoE, complement Cisco’s strategy in supporting cost effective lossless 10 Gigabit Ethernet solutions for the data center.”

-Soni Jiandani, VP of Marketing, Cisco Server Access and Virtualization Group

“Dell’s blade servers are designed to help reduce I/O capital expenses while providing IT and cluster administrators with greater bandwidth performance, better CPU utilization, and more efficient use of their data center space, and Mellanox’s InfiniBand adapter helps us to deliver these gains. Customers with resource-intensive application demands for high-bandwidth and low-latency can leverage blade solutions to improve system efficiency and resource utilization, while still reducing power and overhead costs.”

-Sally Stevens, vice president, product group platform marketing at Dell

“Cloud providers can gain significant competitive advantages by implementing dynamic infrastructures that enable them to improve service, reduce cost and manage risk. InfiniBand simplifies management, and brings better than SAN-like capabilities to every I/O interface, while reducing the number of interfaces, and for this reason, we see more providers building InfiniBand infrastructures today.”

-Jeff Boles, senior analyst and director, validation services at Taneja Group.

“We chose Mellanox InfiniBand in response to customer demand for a scalable Cloud solution as inadequate Gigabit Ethernet performance resulted in customers paying three times as much for the same computing power. By deploying InfiniBand end-to-end, we can offer a high volume and high-performance on-demand service that scales as customers’ needs change and grow. We simply could not be able to achieve this level of performance and flexibility with Ethernet.”

-Dave Durkee, CEO of ENKI.

“The IS5000 switch family with FabricIT fabric management software offers a new level of provisioning flexibility without sacrificing system performance. The scalability of the new switch systems helps data center managers fully optimize their fabrics to meet the demanding requirements in today's dynamic computing environment such as on-demand and cloud computing. Today’s announcement indicates that Mellanox is poised at the forefront of intelligent data center fabric management and performance by continuously rolling out new products to address the emerging market demands.”

-Jie Wu, Research Director, Technical Computing at IDC.

“We seek to increase value for our customers by providing creative finishing tools that grant greater control over final color and lighting effects, as well as improved workflows and efficiencies. The Autodesk pre-configured creative finishing systems help accelerate digital color grading and help collaboration using clustered CPUs connected via Mellanox 40Gb/s InfiniBand or 10 Gigabit Ethernet networking technologies, and provide faster access to data and extensive real-time interaction for demanding client-attended grading sessions.”

-Marc Bolduc, product development manager, Autodesk Media & Entertainment.

“Working with customers like Mellanox and partners like Tela Innovations is the heart of TSMC’s Open Innovation Platform™ initiative. We look forward to continuing this collaboration on future designs.”

-S.T. Juang, senior director, Design Infrastructure Marketing at TSMC.

“When we designed our VMware vSphere-based cloud appliance, large bandwidth, low latency reliability and quick access to storage were of top priorities, and for these reasons InfiniBand became our fabric of choice. Furthermore, InfiniBand also enables a highly secure environment since each VM can reside over a different InfiniBand Virtual LAN, providing strict isolation between VM running in the cloud. It has allowed Intalio to deliver a cutting-edge appliance to our customers that meets their growing demand for an efficient cloud computing infrastructure.”

-Ismael Chang Ghalimi, CEO of Intalio

“With its high-integration, advanced feature-set, and leading performance and power efficiency, the HP NC542m with Mellanox’s ConnectX-2 EN 10 Gigabit Ethernet controller shines as one of the most-attractive 10 Gigabit Ethernet options available through HP. End-users selecting this 10 Gigabit Ethernet adapter can expect to achieve superior data center productivity and service flexibility.”

-Bob Wheeler, senior analyst at The Linley Group

“We were impressed by the performance of the system we tested. On the basis of our test results we would have no hesitation in recommending the InfiniBand connectivity from Mellanox for high-end real-time visualisation with RTT DeltaGen.”

-Christoph Karrasch, co-founder and board member of RTT

“This award, which represents a partnership between the HPCwire global readership and our publishing team, is a salute from the global HPC community. To be selected as an award recipient means you are at the top of the minds of HPCwire readers, editors, and luminaries in the field.  I’d like to congratulate Mellanox for being selected by our editors for both 2009 awards.”

-Tomas Tabor, Publisher of HPCwire

“Collectives Communication is a key limiting factor to parallel application efficiency and scalability. Offloading these communications to the network to minimize the impact of system noise, combined with the use of non-blocking collective operations that enable overlapping communications and computation, provide powerful capabilities for extending application scalability. ORNL, through research on scalable collectives communication sponsored by DOE’s Office of Advanced Computing Research FASTOS program, and Mellanox, have collaborated to find the solution that will enable the next generation of supercomputers to scale more efficiently. The new CORE-Direct technology in ConnectX-2 reflects the results of this joint effort.”

-Richard Graham, distinguished member of the research staff, Computer Science and Mathematics Division, Oak Ridge National Laboratory (ORNL).

“The ASUS Z8PH-D12 SE/QDR is a leading compute-intensive serverboard that is compact, power-efficient and built of high-quality components that reduce power stress while improving durability and product lifespan. By leveraging Mellanox’s low-latency, high-bandwidth 40Gb/s InfiniBand connectivity on our Z8PH-D12 SE/QDR server motherboards, our customers will achieve optimal efficiency and productivity.”

-Tom Lin, general manager of server business unit, ASUS

“The MS-91D1, MS-91D2 & MS-92C2 are our leading compute-intensive server motherboards and systems that feature extremely high-density and energy-efficiency. Mellanox’s low-latency, high-bandwidth 40Gb/s InfiniBand connectivity on our motherboards ensures that our customers can maximize the efficiency and productivity of the MS-91D1 and MS-91D2 processing capabilities.”

- Norman Tsai, VP of Server Business at MSI

“Making the Deloitte Technology Fast 50 is a testament to a company’s commitment to technology. With its 432 percent growth rate over five years, Mellanox Technologies has proven that its leadership has the vision and determination to grow in difficult conditions.”

-Tal Chen, director in charge of the Deloitte Brightman Almagor Zohar Israel Technology Fast 50 Program.

“By working with Mellanox, we will be able to offer our customers efficient, high performing InfiniBand and Ethernet solutions.  Mellanox is a leader in this space, and the company offers cost effective solutions without sacrificing performance or flexibility in the data center.”

-Gary Gammon, senior vice president, enterprise marketing, Bell Micro.

Customers in a broad range of enterprise computing environments, with applications such as database clusters, grid systems, cloud computing and server virtualization are working with ever more complex data sets requiring 10 Gigabit Ethernet bandwidth. Mellanox 10 Gigabit Ethernet adapters provide our customers with a solution that effectively scales and provides the high efficiency required for green and cloud computing.”

-Alex Yost, vice president, IBM Systems & Technology Group.

“With its impressive five-year growth, Mellanox Technologies has earned its position among the fastest growing technology, media, telecommunications, life sciences and clean technology companies in North America.  Deloitte is proud to honor Mellanox Technologies for its achievement.”

-Mark Jensen, Managing Partner, Technology and Venture Capital Services, Deloitte & Touche LLP

“Consistent with its technology leadership track record, Mellanox is pushing the Ethernet technology boundary with the release of its 40 Gigabit Ethernet solution.  IT managers can today utilize Mellanox’s 10 Gigabit Ethernet offerings and migrate to 40 Gigabit Ethernet for their next generation, future proofed data centers. The new ConnectX-2 EN 40G adapter card is an indication of the industry’s insatiable demand for more data center networking performance.”

-Bob Wheeler, senior analyst at The Linley Group

“By meeting the performance, efficiency, scalability and quality of service needs of the most demanding customer and applications, the Intalio|Cloud Appliance, accelerated by Mellanox 40Gb/s InfiniBand, is ideally suited for a broad array of cloud services.  We are pleased to have our joint solution named as one of the ‘Best of VMworld 2009’.”

-Ismael Chang Ghalimi, CEO of Intalio

“This technology allows us to provide greater high-performance computing resources to researchers in our national security programs by simplifying the design, and lowering the cost and power requirements of our scalable units for scientific simulation clusters.  In addition, these new adapters enable higher Lustre file system performance with greater connection flexibility between the InfiniBand cluster interconnect and our 10 Gigabit Ethernet storage area network.”

-Mark Seager, Livermore Computing Asst Dept Head for Advanced Technology, at Lawrence Livermore National Laboratory

"When we designed our first cloud appliance, large bandwidth and low latency were top priorities, and for this reason InfiniBand quickly became a no-brainer.  We selected Mellanox Technologies as exclusive supplier of InfiniBand equipment for the performance and quality of their end-to-end connectivity products, their close relationship with HP, and their world-class technical support. This partnership allowed us to deliver a cutting-edge appliance in a record time, with minimum up-front investments."

-Ismael Chang Ghalimi, CEO of Intalio

“The increasing demands to provide critical business services have expanded the need for higher bandwidth and capabilities. Organizations leveraging the Sun Blade X6275 server module, powered by the new Intel Xeon processor 5500 series and Mellanox’s industry-leading 40Gb/s InfiniBand technology, can maximize the full potential, power, and performance of their systems.”

-Raju Penumatcha, Vice President of Netra Systems and Networking at Sun

“Deployments of virtualization and cloud computing infrastructure are becoming greater in scale and complexity.  Mellanox’s ConnectX-2 adapters, with their combination of high-performance, scalability and protocol flexibility benefits, can help simplify I/O system design and make it easier for IT managers to deploy infrastructure that meets the challenges of a dynamic data center.”

- Cindy Borovick, Research Vice President, Datacenter Networks at IDC

“Customers that use clustering, grid systems and virtualization require technologies that simplify complex data set processing, speeding time to market for scientific and engineering projects.  Mellanox 10 Gigabit Ethernet solutions teamed with HP ProLiant servers enable efficient, scalable customer architectures that deliver significant performance improvements at a lower cost per operation.”

-Steve Cumings, Director of Product Marketing for Scalable Computing and Infrastructure at HP

“As a VMware Ready™ Certified product for VMware vSphere™ 4, the Mellanox 10 Gigabit Ethernet adapter has achieved a significant milestone through its active participation in the VMware Community Source Program.  Mellanox’s 10 Gigabit Ethernet adapters complement the VMware vSphere platform to further enable increased efficiency, control and choice for customers in a converged networking environment.”

-Shekar Ayyar, vice president of infrastructure alliances, VMware

-Christophe Mion, chief technical officer at Scalable Graphics

-Steve Cumings, director of marketing for the Scalable Computing and Infrastructure organization at HP.

-Jie Wu, Research Manager, Technical Computing at IDC.

-Steve Cumings, director of product marketing, Scalable Computing and Infrastructure organization, HP

Aiichiro Inoue, president, next generation technical computing unit at Fujitsu.

-Jag Bolaria, senior analyst at The Linley Group.

-Mark Foss, director of marketing at Arista.

-Kamal Dalmia, vice president of marketing at Teranetics.

-Pete Peterson, enior Vice President and General Manager, Tech Data, Advanced Infrastructure Solutions,

-Gary Gammon, Senior Vice-President of Marketing, Bell Micro

-Jie Wu, Research Manager, Technical Computing at IDC.

-Alex Yost, vice president, System x and BladeCenter, IBM

-Bernie Mills, senior director, alliance programs, VMware.

-Raju Penumatcha, Vice President of Netra Systems and Networking at Sun.

-Sally Stevens, director, PowerEdge Platform Marketing, Dell.

-Jim Pappas, Director of Server Technology Initiatives at Intel.

-Jie Wu, Research Manager, Technical Computing at IDC.

-Shay Hassidim, deputy chief technology officer at GigaSpaces

-Shekar Ayyar, vice president of infrastructure alliances at VMware.

-Raju Penumatcha, Vice President of Netra Systems and Networking at Sun Microsystems.

-Stephen L. Josselyn, Research Director, Global Enterprise Server Solutions at IDC.

-Thomas Lippert, director of the Jülich Supercomputing Centre.

-Vince Mendillo, director of HPC marketing at Microsoft Corp.

-Kevin Kettler, CTO at Dell Inc

-Dave Fellinger, CTO, DataDirect Networks.

-Ed Turkel, product marketing manager, Scalable Computing and Infrastructure Organization, HP.

-Hanxi Chen, Open Storage Product Manager, Sun Microsystems.

-Kamal Dalmia, vice president of marketing at Teranetics. 

-Mark Jensen, Managing Partner, National Venture Capital Services, Deloitte & Touche LLP, in Silicon Valley.

-Juan Loaiza, Oracle Senior Vice President.

-Ed Turkel, manager, product marketing, Scalable Computing and Infrastructure Organization, HP.

-Juan Loaiza, Senior Vice President, Oracle.

-Xavier Cavin, Chief Executive Officer of Scalable Graphics.

-Sanjay Lodha, chief executive officer, Netweb Technologies.

-Jeff Wierer, senior product manager at Microsoft Corp.

-Jim Pappas, Director Technology Initiatives and Industry Marketing at Intel.

-Ed Turkel, manager of HPC product marketing for the Scalable Computing & Infrastructure Organization at HP.

-George Paul, Executive Vice President, HCL Infosystems Ltd.

-Dave Fellinger, CTO of DataDirect Networks.

-Gautam Shah, chief executive officer, Colfax International.

-Eric Stenner, Director of Information Technology at Blue Ridge Numerics

-Jim Ballew, Chief Technology Officer at Appro.

-Scot Schultz, senior strategic alliance manager for HPC, AMD (NYSE: AMD).

-Jie Wu, Research Manager, Technical Computing at IDC.

-Hua Nie, vice president of Dawning Information Industry Co., Ltd

-Nelson Wong, General Manager, at Galactic Computing.

-Sergio Amoni, director of marketing for System x and BladeCenter at IBM.

-Jim Ganthier, director, marketing, BladeSystem, HP

-Tau Leng, GM of HPC and Marketing at Super Micro Computer, Inc. (NASDAQ: SMCI).

-Jim Ballew, CTO of Appro.

-Asaf Somekh, vice president of strategic alliances, Voltaire.

-Bjorn Andersson, director HPC technologies, Sun Microsystems

-Ed Turkel, manager of HPC product marketing for the Scalable Computing & Infrastructure Organization at HP

-John O’Shea, director, Advanced Infrastructure Solutions, at Tech Data.

-Art Wittmann, Editor of InformationWeek.

-Charles Liang, president and CEO of Super Micro Computer, Inc. (NASDAQ: SMCI).

-Scot Schultz, senior strategic alliances manager of high performance computing, AMD

“JPMorgan uses blade servers linked by InfiniBand…and supports 14 applications across multiple lines of business…the firm has seen hardware utilization go from an average of 15 percent of capacity before the project to 70 percent…combined with shrinking unit costs.”

“Data center designers that cannot compromise on latency delays in their applications, and want to consolidate I/O, find InfiniBand’s value proposition to be very compelling. We are happy to be working with Mellanox at several mutual customers to deliver a solution that offers better control over the network fabric and a strong foundation to deploy Xen virtualization for large, multi-node clusters. By using open source InfiniBand software from OpenFabrics.org to add support for InfiniBand based I/O virtualization, we will enable organizations to improve server utilization and create a flexible and responsive data center.”

“InfiniBand building blocks from Mellanox enable compute resources to be dynamically matched with each I/O component, facilitating unprecedented server performance and scalability. The PANTAmatrix system has been able to demonstrate greater than 10X performance improvements for actual compute and I/O intensive customer workloads such as streams processing. In addition, server reconfiguration is possible within minutes, in order to match workload demands, using the System Management interface.”

“InfiniBand’s ability to partition I/O to multiple end-points, and consolidate I/O across data center applications holds the promise for added flexibility and cost savings within VMware environments. Mellanox has a clear commitment to delivering cost-effective virtual infrastructure solutions and has been actively involved in the VMware Community Source program since its inception.  We continue to look forward to working with them in concert with other InfiniBand vendors within the community.”

"A key metric to HPC application scalability is the bandwidth balance between the system processor, memory, PCI Express, and the cluster network. The recent introduction of dual-core processors places additional data contention on the servers I/O system. The introduction of DDR solutions ensures that InfiniBand will remove the network bottleneck in these next generation servers."

"Linux Networx focuses on delivering new, innovative technologies that improve the performance of our Evolocity* computing systems. The software compatibility between 10Gb/s and 20Gb/s InfiniBand technology makes it possible for our customers to recognize the performance benefits of DDR technology. Our customers are interested in economically scaling from small configurations to multi-thousand node supercomputing environments, and DDR InfiniBand is an attractive choice for bringing high-performance compute and storage management together."

"DDR technology further increases the strategic importance of using InfiniBand to directly attach servers to storage through a unified fabric. A 20Gb/s InfiniBand connection provides the highest performance available on a single industry standard I/O Storage link, ultimately delivering world-class performance for high-throughput HPC storage applications."

"InfiniBand has enabled the Isilon IQ clustered storage system to easily scale past 150 terrabytes of capacity with 3 gigabytes per second of storage performance from a single file system. We picked InfiniBand with its aggressive roadmap to 20Gb/s DDR technology and beyond as it gives us a competitive edge over all other interconnect solutions and allows us to scale our clustered storage solutions for many generations to come. We applaud Mellanox for driving this technology into the marketplace."

"InfiniBand DDR allows Voltaire to extend its leadership position in the area of high performance grid computing. As clusters continue to grow in size, demand has increased for bandwidth as well as support for higher aggregate bandwidth per switch. DDR solutions allow these objectives to be met and continue to increase the overall performance of compute clusters and grids."

"Lawrence Livermore National Laboratory's high-performance computing cluster environments are expanding rapidly from hundreds of nodes to multiple thousands of nodes. The 60Gb/s uplinks for inter-switch communication provide a cost effective solution for large clusters and create a more manageable cable infrastructure as only a single 12X cable is required. We are encouraged by the advance of OpenIB in unifying the InfiniBand marketplace and in making it more accessible. Taken together, these innovations advance high performance scientific computing by allowing research programs, such as those at Livermore, to build large clusters out of four socket nodes."