HP announces OpenFlow-enabled switches.  HP (HPQ) announced a portfolio of OpenFlow-enabled switches, providing customers with the broadest choice in the industry for simplifying network management while meeting a wide range of bandwidth, performance and budget needs. The sixteen model portfolio includes the HP 3500, 5400 and 8200 series switches.  Later this year HP will have support for OpenFlow across all switches in the HP FlexNetwork architecture. “To help enterprises uniformly manage network traffic flow across the data center, HP became an early innovator of the OpenFlow standard, supporting more than 60 universities and research centers in trials, development efforts and practical real-world applications,” said Bethany Mayer, senior vice president and general manager, Networking, HP. “With more than 10 million OpenFlow-capable switch ports deployed and the broadest portfolio currently available on the market, HP is leading the move to the OpenFlow standard and further simplifying networks for our enterprise clients.” HP is a founding member of the Open Networking Foundation which continues to advance the OpenFlow standard.

$200,000 Cray Supercomputers.  Cray announced that it is revitalizing its efforts to provide affordably priced, world-class Cray supercomputers to customers in the midrange supercomputing market. With prices starting at $200,000, new entry-level midrange configurations combine the lower cost with a range of software application support previously reserved for the much larger petascale technologies. The new offering features enhancements to Cray’s Cluster Compatibility Mode (CCM), which gives customers the ability to run applications from Independent Software Vendors (ISVs) without modification. ”The Cray CX line was a success for us as it allowed us to reach a new segment of users in a broader set of industries,” said Peg Williams, Cray’s senior vice president of high performance computing systems. “For some time, customers have been looking for us to combine the strengths of both of our product offerings into a single architecture and we’ve now accomplished that goal. By adding the final pieces to the puzzle — low starting prices, broad ISV applications support, scalability and a tightly integrated architecture – we now have an ideal supercomputing product to successfully meet the needs of the midrange market.”

UPSL expands PowerWAVE 6000 UPS.  UPS Limited, a Kohler company announced that the PowerWAVE 6000 UPS is now available in both 400 and 500 kVA power ratings, parallelable up to 5 MVA. The new UPS was specifically designed to meet the growing power protection demands of mission critical applications within the data centre, IT, telecommunications and financial services sectors. “We have seen demand for the existing PowerWAVE 6000 continue to grow, especially within the small to medium sized data centre arena, where space and efficiency are key considerations,” said David Renton, UPSL’s Managing Director.  ”However, feedback from both existing and potential customers, regularly asked if there was a UPS which maintains the same advanced characteristics, but with a larger kVA rating. In response, we extended the range to produce a UPS that met the required power capacity but without impacting the unit’s performance or efficiency credentials. This class-leading UPS system is now available in ten power ratings, ranging from 60 to 500 kVA, and boasts an output power factor rated up to one – ensuring compliance with all modern IT equipment like blade servers.”

Intel to acquire Qlogic’s InfiniBand business. Intel Corporation (INTC) announced that it has entered into a definitive agreement with QLogic to acquire the product lines of and certain assets related to its InfiniBand business. The InfiniBand assets help Intel grow its networking portfolio and provide scalable high performance computing (HPC) fabric technology. InfiniBand is widely used in HPC environments and is a key element to Intel’s efforts to innovate fabric architectures in order to achieve ExaFLOP per second performance by 2018.  InfiniBand was the interconnect family used in 41.8 percent of the Top500 Supercomputers in November 2011. “At the International Supercomputing Conference 2011, Intel unveiled a bold vision to redefine HPC performance and break the Exascale barrier by 2018,” said Kirk Skaugen, vice president and general manager of Intel’s Data Center and Connected System Group. “The technology and expertise from QLogic provide important assets to provide the scalable system fabric needed to execute on this vision. Adding QLogic’s InfiniBand product line to our networking portfolio will bring increased options and exceptional value to our datacenter customers.”

New release of Rackwise DCIM product. Rackwise announced that it will release version 3.6 of its core DCIM software product, the Rackwise Data Center Manager in April of 2012.  The new version will be released in accordance with its recent licensing agreement with Intel. The new software release will contain a suite of new features including enhanced device level real-time monitoring, enhanced power modeling, rich asset search, visualization, blade server management, fault impact analysis and business reporting analytics. “We are extremely excited to make this announcement, marking the initial product launch pursuant to our agreements with Intel to develop and introduce ground-breaking capabilities for the data center infrastructure management market,” said Rackwise CEO Guy Archbold. “The integration of device level data aggregation within our asset visualization, modeling, and analytics functions provides data center and infrastructure technology professionals powerful, state-of-the-art tools for real-time data management to maximize operational efficiencies and economic benefits.”

Fibertown adds carriers to Houston data center.  Fibertown announced the addition of new telecom carriers into its 50,000 sq.ft. North Houston data center. “Multi-carrier connectivity provides our data center customers with more robust options and plays a crucial role in the redundancy of our offerings,” said Brandon Perryman, senior vice president at FIBERTOWN. “Customers are empowered to select acarrier that best fits their business requirements.”  The Houston data center offers diverse fiber entrances and houses primary and secondary Point of Presence (PoP) rooms.

Internap completes SOC2 audit.  Internap (INAP) announced that it has successfully completed the transition from SAS 70 to a Service Organization Controls (SOC) Report, regarding the operations and control structure in its company-controlled data centers. The SOC2 Type II Report confirms that Internap’s data center security and operational procedures have been reviewed and tested by an independent certified auditor and validates that controls and processes are suitably designed and operate effectively to protect and safeguard customers’ equipment and data. “In today’s global economy, service providers must demonstrate they have adequate controls and safeguards when they host information belonging to their customers,” said Mike Higgins, senior vice president of data center services at Internap. ”Completing the SOC2 audit demonstrates Internap’s commitment to meeting the service and performance level needs of our customers.”

An exterior view of the National Petascale Computing Facility where the Blue Waters Supercomputer will reside.

The National Center for Supercomputing Applications (NCSA) announced that the National Petascale Computing Facility (NPCF) at the University of Illinois has been earned a Gold-level certification under the LEED (Leadership in Energy and Environmental Design) rating program for energy-efficient buildings.

Blue Waters

Constructed in 2010 the University of Illinois and NCSA opened the NPCF data center as the home to supercomputers and other high-performance systems operated by NCSA and used by scientists and engineers across the country. The Blue Waters project encompassed the NPCF and a 10 petaflop supercomputer, which was initially a venture with IBM. In 2011 NCSA and IBM determined that the project was too complex to proceed. IBM pulled the plug and NCSA later awarded they contract to Cray to build a XE6 system.

The features of the NPCF data center that led to a LEED Gold certification include a 480V power distribution system for computating equipment, a focus on water-cooled computational and storage equipment, external cooling towers for free cooling, low-impact landscaping with native prairie plants and using best practice construction methods to improve the air quality environment within the facility.

LEED Gold was the target established back in 2010 by the University of Illinois, along with a pledge to take steps toward carbon neutrality, reduced energy use and overall improved sustainability in the future. The NPCF receives chilled water from the University chiller plant and also has outside thermal storage tanks that are cooled approximately 70 percent of the year by mother nature. This is expected to cut the chilled water costs for the facility by about $1 million per year.

Cray delivers first Blue Waters Cabinet. On December 1 Cray delivered the first full cabinet for the NCSA Blue Waters system. A photo gallery of the installation day can be found on the NCSA Facebook album, where in the comments it is confirmed that the cabinets will be water-cooled. The National Science Foundation’s Blue Waters project was awarded to Cray last month after NCSA and IBM terminated the original contract last summer.

Dell’s HPC Strategy.  The Register reports on how Dell is going to engage the market to grow its HPC strategy. The primary focus for Dell’s HPC strategy is to concentrate on smaller HPC systems where projects are well-bounded with known workloads and customers they know and understand. Dell is putting together recipes for popular HPC apps in small, medium and large configurations. Each of these is fully tested and guaranteed to hit the promised level of performance. Dell’s  High Performance Computing web site lists strategic insights, recommended configurations and a High Performance Cluster Advisory tool.

SGI selected for Poznan HPC System.  SGI announced that Poland’s Poznan Supercomputing and Networking Center (PSNC) has purchased an SGI HPC solution. The purchased system is comprised of an SGI Rackable C1103-G15 cluster utilizing AMD Opteron 6200 series processors, and includes 120 servers each with one NVIDIA M2050 GPU per node and another 107 servers with two M2050 GPUs per server. The full configuration is expected to deliver 224 peak teraflops and will feature 12.6 TB of memory, 5448 CPU cores and 149632 GPU CUDA cores, and is intended to satisfy the increasing performance requirements of Poznan Supercomputing Center users. ”We decided to go for the combination of the most modern CPU technology together with proven GPU technology,” said Norbert Meyer, director of the Supercomputing facility at Poznan. “Such a combination allows us to ensure optimal parameters for our users. This system, which was financed from structural funds coming from POWIEW and PL-GRID national projects and will be also used in the PRACE (EU project), allowed us to reach 63 teraflops, enough for a place on the TOP500 list today, with less than half of nodes installed for the test. The full configuration will feature more than twice as many more nodes than the measured system, which ensures enough performance for our users as well as a good position on June 2012 TOP500 list.”

The twice-a-year list of the Top 500 supercomputers documents the most powerful systems on the planet. Many of these supercomputers are striking not just for their processing power, but for their design and appearance as well. Here’s a look at the top finishers in the latest Top 500 list, which was released Monday at the SC11 conference in Seattle.

IBM Announces Blue Gene/Q.  IBM announced its next generation supercomputing project, Blue Gene/Q, will provide an ultra-scale technical computing platform to solve the most challenging problems facing engineers and scientists at faster, more energy efficient, and more reliable rates than ever before. When it is fully deployed in 2012 at Lawrence Livermore National Laboratory (LLNL), the system, named “Sequoia”, is expected to achieve 20 petaflops at peak performance. IBM and LLNL are also aiming for extreme energy efficiency by achieving 2 gigaflops per watt with Blue Gene/Q.  Blue Gene/Q will be used to help predict the path of hurricanes, analyze the ocean floor to discover oil, simulate nuclear weapons performance and decode gene sequences. “Completing computationally intensive projects for a wide variety of scientific applications that were previously unsolvable is not just possible – it is now probable,” said Brian Connors, VP of technical computing at IBM. “IBM’s historic role in developing the supercomputers that provide the power behind critical applications across every industry has uniquely positioned us to provide reliable supercomputing at the highest level.”

Cray enters storage market and touts OpenACC.  Cray announced the latest addition to its line of high performance computing (HPC) products with the launch of the Cray Sonexion 1300 system — an integrated file system, software and storage product. With this Cray can offer a complete and scalable Lustre file system that will scale from 50 terabytes to more than 50 petabytes of useable capacity. The Cray Sonexion 1300 storage system will be a key part of the recently announced Blue Waters supercomputer at the National Center for Supercomputing Applications at the University of Illinois, and will provide more than one terabyte-per-second of aggregate bandwidth. ”More and more of our customers have expressed a need for increased performance, reliability and manageability from storage systems that can be shared with their highest performance, mission critical systems in their datacenter,” said Barry Bolding, Cray’s vice president of storage and data management. “As a result, we decided to provide a storage product with a scale-out architecture, which is purpose-designed and built to run Lustre, but employs a testing discipline and set of integrated management tools that exceeds other available solutions. The Cray Sonexion 1300 storage solution greatly improves the user experience with shared, scalable Lustre file systems – both large and small.”  Cray also joined NVIDIA, the Portland Group (PGI), and CAPS enterprise to announce a new parallel-programming standard, known as OpenACC. OpenACC is a new open parallel programming standard designed to enable the millions of scientific and technical programmers to easily take advantage of the transformative power of heterogeneous CPU/GPU computing systems.

Extreme Networks and Schneider power SC11.  Bringing together  infrastructure behind the scenes for a conference about high-performance computing is a chance for vendors to show off their high-performance gear.  Extreme Networks (EXTR) is providing Ethernet switching solutions addressing the ultra-high speed, energy efficiency and latency needs. The company this week is showing its recently tested BlackDiamond X8 switch, which was profiled as 3-10 times faster than any core switch in the market in this Lippis Report video. At the company’s conference booth they will be discussing how its 10/40GbE Open Fabric switching network solutions, the BlackDiamond X8 and Summit X670 switches, featuring the ExtremeXOS modular operating system, meet the needs of HPCC and emerging Big Data Hadoop clusters. Schneider Electric was named as an official provider of power for supporting the conference. During the show high performance applications will be running through the data center backed by APC by Schneider Electric’s Symmetera PX 250 UPS and InfraStruxure 150kW PDU.

SC11 updates can be found through the Twitter hashtag #SC11 or @Supercomputing

An exterior view of the National Petascale Computing Facility where the Blue Waters Supercomputer was scheduled to reside.

NCSA and Cray announced that they have finalized a contract with the University of Illinois’ National Center for Supercomputing Applications (NCSA) to provide the supercomputer for the National Science Foundation’s Blue Waters project. Back in August NCSA and IBM jointly announced that IBM has terminated its contract with the University of Illinois.

The Blue Waters Infrastructure

The multi-phase, multi-year project was awarded to Cray for $188 million and will start with a Cray XE6 system, upgrading to the recently announced Cray XK6 with built-in GPU computing capability.  Bill Kramer, deputy project director of the Blue Waters project at the NCSA at the University of Illinois, told The Register that Blue Waters was not a specific system, but rather a complete set of infrastructure, including a data center, plus computation, networking, and storage and, most importantly given the software goals of the NCSA, code that scales to real-world petaflops performance.

The completed system will support significant research advances in a broad range of science and engineering domains, meeting the needs of the most compute-intensive, memory-intensive, and data-intensive applications. ”The project is an incredible undertaking, requiring commitment and dedication not only from NSF, NCSA, the University of Illinois, and the science teams, but also from our computing systems partner—Cray,” said University of Illinois President Michael Hogan. ”This strong partnership further establishes our place at the forefront of high-performance computing.”

“We’re very excited to have been selected by the NCSA, NSF and the University of Illinois to deliver the Blue Waters system, which represents one of the largest contracts in our company’s history,” said Peter Ungaro, president and CEO of Cray. “Together with the recently announced $97 million contract to upgrade the ‘Jaguar’ system at Oak Ridge National Laboratories, these contracts demonstrate Cray’s leadership position in supercomputing. With a strong core business and future growth opportunities around our new initiatives, we are extremely pleased to be able to provide our outlook for 2012 of strong revenue growth and profitability.”

Cray and NVIDIA

The completed system will feature new 16-core AMD Opteron 6200 Series processors, a next-generation graphics processing unit (GPU) from NVIDIA called Kepler, and a new integrated storage solution from Cray.  The top Cray system on the Top 500 list, Jaguar, will also rely heavily on NVIDIA GPU’s and a mix of XT5 and XT6 cabinets as it upgrades to be known as Titan.

The new Blue Waters system will have at least 235 XE6 cabinets, more than 30 XK6 cabinets, and more than 30 storage and I/O server cabinets. The resulting machine will have more than 49,000 Opteron 6200 processors and more than 380,000 cores, with another 3,000 Nvidia GPU coprocessors. It will also take advantage of Cray’s Gemini high-performance interconnect, providing a major improvement in message throughput and latency. Blue Waters will run the Cray Linux Environment and use the storage subsystem under development at Cray ,running the Lustre file system. It is expected to deliver 25 petabytes of usable disk capacity.

Acceptance of the complete system is anticipated to occur in the fourth quarter of 2012.

The K Computer from Japan, shown above, has retained its title as the world's most powerful supercomputer in the new Top 500 rankings released earlier today.

The newly-released Top 500 list of the world’s most powerful supercomputers finds Japan’s K Computer defending its number one spot, increasing from 8.162 petaflops to 10.51 petaflops per second (10.51 quadrillion floating-point operations per second). The letter “K” is short for the Japanese word “kei,” which symbolizes 10 quadrillion.

A Familiar List

On the June Top 500 list, the K computer took the number one spot away from China’s China’s Tianhe-1A system, which remains in the number two spot at 2.57 petaflops.  With the Oak Ridge National Lab’ss Jaguar supercomputer at number three, the rest of the 38th edition of Top 500 list looks pretty familiar. “This is the first time since we began publishing the list back in 1993 that the top 10 systems showed no turnover,” said TOP500 editor Erich Strohmaier, who will lead the discussion at SC11.

10 Petaflops on the way to 23

Installed at the RIKEN Advanced Institute for Computational Science (AICS) in Kobe, Japan in partnership with Fujitsu, the K Computer uses 705,024 SPARC64 processing cores, which is more than the rest of the top five on the list combined.  Four times more powerful than its nearest competitor, the K Computer uses 12.66 megawatts of power, compared with 9.89 megawatts reported in June.  It is however still one of the most efficient supercomputers on the list though, delivering 830 Mflops/watt.

The K Computer was installed in 672 racks in June and an expansion to 800 cabinets allowed the 10 petaflop achievement to be accomplished. A week ago Fujitsu announced the expansion possibilities for the K computer as it could grow to a theoretical 23 petaflops.

Power and Performance

Although the rankings are unchanged, other characteristics evolved with the November 2011 list.  The entry point for being on the Top 500 list is now 50.9 teraflops, and the combined performance across all 500 superomputers is 74.2 petaflops, up from 58.7 petaflops last June. Used in 223 systems, Gigabit Ethernet is still the most-used internal system interconnect, but InfiniBand use increased to 213 systems. As the average power consumption continued to rise, 29 of the systems on the list are now confirmed to use more than one megawatt of power. The most energy efficient supercomputers are BlueGene/Q with 2,029 Mflops/watt. IBM recently filed a patent for a massive supercomputer that could reach 107 petaflops. Of those systems reporting on power the total consumption of the 500 systems is more than 159 megawatts.

“The road to Exascale” is a frequent article title when talking about the future of supercomputer systems and with that goal in mind, large efficiency gains will be front and center for manufacturers, to avoid needing gigawatts of power.

Data Direct Networks introduces big data appliances.  Data Direct Networks (DDN) launched the SFA 12K series of Big Data appliances, a suite of products designed to accelerate data-intensive applications and consolidate infrastructure to minimize total cost of data storage ownership for Big Data, high-performance computing (HPC), cloud and content-intensive environments.  Three new models of DDN’s Storage Fusion Architecture accelerate application processing and enhance data center efficiency.  The SFA 12K-40, when combined with InfiniBand and Fibre Channel connectivity the new appliance will scale file storage systems to 1 Terabyte per second. The SFA 12K-20E features highly-optimized virtualization and x86 co-processing, and the SFA 12K-20 combines best-in-class storage performance with a cost-optimized system configuration to economically deliver over 20 Gigabytes per second and 700,000 Flash IOPS from a single system. “The exponential growth of Big Data is putting significant strain on IT infrastructures and challenging organizations’ ability to successfully ingest, store, process and distribute data,” said Alex Bouzari, CEO and cofounder, DataDirect Networks. “With the SFA12K series’ groundbreaking ability to virtualize applications and store and process within the same appliance, we are delivering ideal storage and compute infrastructure solutions for organizations worldwide.” InsideHPC interviews DDN’s Jeff Denworth to introduce the SFA12K.  SFA12K systems will ship in the second quarter of 2012.

Virident launches high performance memory.  Virident Systems announced the release of Virident FlashMAX MLC, a PCIe storage class memory (SCM) solution that offers enterprises unconditional performance and the industry’s highest storage capacity at 1.4 TB in a low-profile form factor. With this addition to the company portfolio it is able to address the growing needs of enterprise for high performance, high capacity flash storage solutions in their data centers. ”Flash storage performance varies by type of course, but it is also considerably affected by factors such as application workloads and data sets, and will invariably downgrade as the capacity of the disk gets used,” said Mark Peters, senior analyst at the Enterprise Strategy Group. “Virident’s FlashMAX SCM solution, integrated with vFAS software, has been specifically designed to eliminate these variations and deliver consistent performance over time – something that is critical in order for users to gain the maximum benefits from their investments in flash storage.” Virident also announced that it has closed a $21 million series C funding round with strategic investments from Intel Capital, Cisco and others, bringing total funding up to $37 million.

Intel plans for Sandy Bridge HPC boards.  CRN.com reports that next week at the SC11 conference Intel (INTC) plans to mark its re-entry into the high performance computer market with the introduction of new motherboards and chasses its system builder and integrator partners can use to develop HPC offerings. Intel’s director of marketing for its Enterprise Platform and Services Division (EPSD) told CRN.com that Intel will use SC11 to introduce a new series of half-width, dual-socket motherboards with high-performance throughput and large memory that take advantage of the company’s release in the first half of 2012 of its new Romley architecture with Sandy Bridge processors.

Ciena 100G innovations at SC11.   Ciena (CIEN) announced the details of its advanced optical networking demonstrations at the upcoming SuperComputer Conference 2011 (SC11) in Seattle. Additionally, for the fourth year in a row Ciena will provide 100G optical infrastructure for SCinet, the high-performance network built for the SC11 conference. Ciena demonstrations can be found at Internet2 and ESnet (booth #1327 and #512), NASA/ICAIR (booth #615 and 2615), BCNET, CANARIE, Caltech, University of Victoria (booth #3445) and the University of Amsterdam (Ciena booth #635). Ciena executives will also be presenting at the conference.

Luxtera delivers 100Gbps Opto-Electronic Transceiver.  Luxtera announced the industry’s first single chip 100Gbps optical transceiver to support next generation cloud computing data centers and high performance computing (HPC) optical connectivity. The single chip opto-electronic transceiver is targeted at applications that would require 100 Gbps Ethernet, OTN or InfiniBand. The fully integrated chip leverages silicon photonics and uses light from a single co-packaged laser to power multiple optical transmitters on a chip, eliminating the need for multiple lasers and reducing transceiver cost and power consumption. “The introduction of the 100Gbps Silicon Photonics transceiver, which is the outcome of a joint Luxtera-Molex collaboration, is a key milestone in optical connectivity for a wide range of cloud computing, data center and HPC applications.  We are excited about this collaboration with Luxtera as it represents one of the many positive outcomes of our partnership and recent agreements,” said Doug Busch, Vice-President and General Manager of Molex’s Fiber Optic Products Business Unit. “As a strategic partner with Luxtera, Molex will deliver a line of connectivity products based on this IP spanning different data rates, lanes and mechanical form-factors.”

NASA upgrades with SGI ICE HPC.  SGI announced today that NASA has selected its next generation SGI ICE high performance computing (HPC) platform, code-named ‘Carlsbad 3.0,’ to extend the computational capability of NASA’s Pleiades supercomputer system. Starting in 2012 NASA will expand their computational capabilities with the new SGI platform, future Intel Xeon E5 Romley family processors, FDR dual-plane and a hypercube-topology InfiniBand network. With the goal of attaining 10 petaflops of peak performance NASA has partnered with SGI and Intel under the Space Act Agreement for research, modeling and simulation work at the NASA Advanced Supercomputing (NAS) facility at NASA’s Ames Research Center at Moffett Field, California. The new system will be integrated into the Pleiades supercomputer, ranked number 7 on the June 2011 Top500 list. ”Intel and SGI are working closely together to make sure that leading customers like NASA have the performance, density and programming efficiency they need to deliver on key multi-petascale workloads leading up to the exascale era,” said Rajeeb Hazra, general manager of the High Performance Computing division at Intel. “We’re excited that NASA selected the fifth generation of SGI’s award-winning ICE system for the next stages of Pleiades expansion.”

NetApp and Cloudera partner.  NetApp (NTAP)  announced a preconfigured, ready-to-deploy solution called NetApp Open Solution for Hadoop that will enable customers to maximize the value of their enterprise Hadoop implementations with improved flexibility and performance and lower total cost of ownership. Partnering with Cloudera they will distribute Cloudera’s Distribution including Apache Hadoop (CDH) and Cloudera Enterprise, a subscription service comprised of Cloudera Support and Hadoop management software, with NetApp Open Solution for Hadoop to speed enterprise deployment and production use of Apache Hadoop. ”NetApp has a long history of supporting open standards and giving our customers the flexibility and efficiency they need to drive results from their data. Our strategy for Hadoop is no different, and we will give customers choice for their Hadoop implementations, from enterprise-class to open-source Apache Hadoop solutions,” said Clifton. “Cloudera strengthens our enterprise-ready NetApp Open Solution for Hadoop, and together we make deployment of enterprise Hadoop easier, simplify management, and reduce the cost of deployments.” Hadoop World was a sold out event November 8-9 in New York.