HPC Wire

Subscribe to HPC Wire feed
Since 1987 - Covering the Fastest Computers in the World and the People Who Run Them
Updated: 20 hours 42 min ago

Oklahoma State University Selects Advanced Clustering to Build New HPC Cluster

Mon, 11/06/2017 - 09:26

KANSAS CITY, Mo., Nov. 6, 2017 — Oklahoma State University has selected Advanced Clustering Technologies to build and install its newest supercomputer, which will support computing- and data-intensive research across a broad range of Science, Technology, Engineering and Mathematics (STEM) disciplines.

The new supercomputer, which will be named after the university’s mascot, Pistol Pete, will serve as a campus-wide shared resource, available at no charge to all OSU faculty, staff, postdocs, graduate students and undergraduates, as well as to researchers and educators across Oklahoma.

“Pistol Pete will make it possible for us to continue to meet the ever-growing demand for computational and data-intensive research and education at OSU and across Oklahoma,” said Dr. Dana Brunson, Assistant Vice President for Research Cyberinfrastructure and Director of the OSU High Performance Computing Center as well as an Adjunct Associate Professor in the Mathematics and Computer Science departments. “This new cluster allows us to extend the mission of our High Performance Computing Center, which is to put advanced technology in the hands of our faculty, staff and students more quickly and less expensively, and with greater certainty of success.”

Dr. Brunson is principal investigator of the grant proposal, Acquisition of Shared High Performance Compute Cluster for Multidisciplinary Computational and Data-Intensive Research, that was awarded a $951,570 Major Research Instrumentation grant by the National Science Foundation.

The new cluster is powered by Intel® Xeon® Scalable Processors (previously known as Skylake). Pistol Pete consists of more than 5,300 compute cores in Advanced Clustering’s innovative ACTblade x110 systems. Along with the large compute capacity, a high-speed Lustre storage system and low-latency 100Gb/s Omni-Path fabric is part of the turn-key cluster being delivered.

“We are proud to be working with Dr. Brunson and Oklahoma State University to provide this new high performance computing resource,” said Kyle Sheumaker, President of Advanced Clustering Technologies, which specializes in providing turn-key HPC resources to universities, government agencies and enterprises across the country. Advanced Clustering also built the university’s previous cluster, called Cowboy, which at its deployment in 2012 was a 3,048-core cluster using servers based on the Intel Xeon processor E5 family. Cowboy gave the center nine times the capacity of its previous cluster in only two times the physical space.

Pistol Pete will support research in more than 30 departments and many collaborating research teams, including 129 faculty and more than 740 postdocs, graduate students and undergraduates. Research topics include bioinformatics, biomolecular model development; crop modeling, environment and ecosystem modeling; classical and quantum calculations of liquids, proteins, interfaces and reactions; design and discovery of organic semiconductor materials; cybersecurity and social network modeling; commutative algebra; graph-based data mining; renewable energy research; seismology; sociopolitical landscape modeling; and high energy and medical physics.

About the High Performance Computing Center at Oklahoma State University
The High Performance Computing Center (HPCC) facilitates computational and data-intensive research across a wide variety of disciplines. HPCC provides Oklahoma State University students, faculty and staff with cyberinfrastructure resources, cloud services, education and training, bioinformatics assistance, proposal support and collaboration. Learn more about HPCC resources at https://hpcc.okstate.edu/.

About Advanced Clustering Technologies 

Advanced Clustering Technologies has been building customized, turn-key high performance computing clusters, servers and workstations since the company’s founding in 2001. The company specializes in developing campus-wide supercomputers, which when fully utilized can help maintain and grow a university’s competitive edge by attracting researchers and students from around the globe. Advanced Clustering’s knowledgeable team of HPC experts provide phone and email support for the lifetime of your system. For more information, visit advancedclustering.com.

Source: Advanced Clustering Technologies

The post Oklahoma State University Selects Advanced Clustering to Build New HPC Cluster appeared first on HPCwire.

Mellanox Announces Innova-2 FPGA-Based Programmable Adapter Family

Mon, 11/06/2017 - 07:45

SUNNYVALE, Calif. & YOKNEAM, Israel, Nov. 6, 2017 — Mellanox Technologies, Ltd. (NASDAQ: MLNX), a leading supplier of high-performance, end-to-end smart interconnect solutions for data center servers and storage systems, today announced the Innova-2 product family of FPGA-based smart network adapters. Innova-2 is the industry leading programmable adapter designed for a wide range of applications, including security, cloud, Big Data, deep learning, NFV and high performance computing.

Innova-2 is available in multiple configurations, either open for customers’ specific applications or pre-programmed for security applications with encryption acceleration such as IPsec, TLS/SSL and more. For security applications, Innova-2 delivers 6X higher performance while reducing total cost of ownership by 10X when compared to alternative options. For Cloud infrastructures, Innova-2 enables SDN and virtualized acceleration and offloads. Deep learning training and inferencing applications will be able to achieve higher performance and better system utilization by offloading algorithms into Innova-2 FPGA and the ConnectX acceleration engines.

Innova-2 is based on an efficient combination of the state-of-the-art ConnectX-5 25/40/50/100Gb/s Ethernet and InfiniBand network adapter with Xilinx UltraScale FPGA accelerator. Innova-2 adapters deliver best-of-breed network and storage capabilities as well as hardware offloads to CPU-intensive applications.

“The Innova-2 product line brings new levels of acceleration to Mellanox intelligent interconnect solutions,” said Gilad Shainer vice president of Marketing, Mellanox Technologies. “We are pleased to equip our customers with new capabilities to develop their own innovative ideas, whether related to security, big-data analytics, deep learning training and inferencing, cloud and other applications. The solution allows our customers to achieve unprecedented performance and flexibility for the most demanding market needs.”

The Innova-2 family of dual-port Ethernet and InfiniBand network adapters supports network speeds of 10, 25, 40, 50 and 100Gb/s, while the PCIe Gen4 and OpenCAPI (Coherent Accelerator Processor Interface) host connections offer low-latency and high-bandwidth. Innova-2 allows flexible usage models, with transparent accelerations using Bump-in-the-Wire or Look-Aside architectures. The solution fits any server with its standard PCIe card form factor (Half Height, Half Length), enabling a wide variety of deployments in modern data centers.

“We are happy to continue our deep collaboration efforts with Mellanox and to deliver optimized compute platforms to our mutual customers,” said Brad McCredie, vice president, Cognitive Systems Development, IBM. “Innova-2 will enable us to maximize the performance of our cloud, deep learning and other platforms by utilizing Mellanox acceleration engines and the ability to connect to our industry leading OpenCAPI CPU interfaces.”

“At Secunet Security Networks AG, we specialize in bringing high-performance security solutions to our very demanding government and commercial customers,” said Dr. Kai Martius, CTO of Secunet. “The Innova family of adapters is an ideal element to our network security products, delivering protocol-sensitive accelerated crypto processing to our offerings.”

“Xilinx is pleased that our All Programmable UltraScale FPGAs are accelerating Mellanox’s Innova network adaptors,” said Manish Muthal, vice president of Data Center Business at Xilinx. “Our combined technology enables the rapid deployment of customized acceleration for emerging data center and high performance computing workloads.”

Mellanox will be showcasing Innova-2 adapters at the OpenStack Summit, Nov 6-8, booth B20, Sydney, Australia, and at SC17, Nov 13-16, booth # 653, at the Colorado Convention Center.

About Mellanox

Mellanox Technologies (NASDAQ: MLNX) is a leading supplier of end-to-end InfiniBand and Ethernet smart interconnect solutions and services for servers and storage. Mellanox interconnect solutions increase data center efficiency by providing the highest throughput and lowest latency, delivering data faster to applications and unlocking system performance capability. Mellanox offers a choice of fast interconnect products: adapters, switches, software and silicon that accelerate application runtime and maximize business results for a wide range of markets including high performance computing, enterprise data centers, Web 2.0, cloud, storage and financial services. More information is available at: www.mellanox.com.

Source: Mellanox

The post Mellanox Announces Innova-2 FPGA-Based Programmable Adapter Family appeared first on HPCwire.

Trends in HPC and Machine Learning Drive Evolution of Cooling Solutions

Mon, 11/06/2017 - 01:01

As seen at ISC17 and will be seen at SC17, the application of HPC in finance, logistics, manufacturing, big science and oil & gas is continuing to expand into areas of traditional enterprise computing often tied to the exploitation of Big Data. It is clear that all of these segments are using (or planning to use) machine learning and AI resulting in architectures that is very HPC-like.

The physical implementation of these systems requires a greater focus on heat capture and rejection due to the wattage trends in CPUs, GPUs and emerging neural chips required to meet accelerating computational demands in HPC-style clusters. The resulting heat and its impact on node, rack and cluster heat density is seen with Intel’s Knight’s Landing and Knight’s Mill, Nividia’s P100 and the Platinum versions of the latest Intel Skylake processors.

Wattages are now sufficiently high that to cool nodes containing these highest performance chips leaves one with little choice other than liquid cooling to maintain reasonable rack densities. In sustained compute sessions, there must be no throttling or down-clocking of the compute resources. If not addressed at the node level with liquid cooling, floor space build-outs or data center expansions become necessary. Even more importantly, reducing node and rack densities can drive an increase in interconnect distances between all types of cluster nodes.


Asetek RackCDU D2C™ Cooling

These developments are a direct result of a wattage inflection and not simply an extension of trends seen previously. Depending on the approach taken, machine learning and AI exacerbate this trend. Heat and wattage issues seen with GPUs during the training or learning phase of an AI application (especially if used in a deep learning/ neural network approach) are now well known.  And in some cases, these issues continue into application rollout if GPUs are applied to that as well.

Even if the architecture uses quasi-GPUs like Knight’s Mill in the training phase (via “basic” machine learning or deep learning followed by a handoff to scale-out CPUs like Skylake for actual usage) the issues of wattage/density/cooling remains. And it isn’t getting any better.

With distributed cooling’s ability to address site needs in a variety of heat rejection scenarios, it can be argued that the compute-wattage-inflection-point is a major driver in the accelerating global adoption of Asetek liquid cooling at HPC sites and by the OEMs that serve them.  And as will be shown at SC17, quite of few of the nodes OEMs are showing with liquid cooling are targeted at machine learning.

Given the variety of clusters (especially with the entrance of AI), the adaptability of the cooling approach becomes quite important. Asetek distributed pumping architecture is based on low pressure, redundant pumps and closed loop liquid cooling within each server node. This allows for a high level of flexibility in heat capture and heat rejection.

Asetek ServerLSL™ is a server-level liquid assisted air cooling (LAAC) solution. It can be used as a transitional stage in the introduction of liquid cooling or as a tool to enable the immediate incorporation of the highest performance computing nodes into the data center. ServerLSL allows the site to leverage existing HVAC, CRAC and CRAH units with no changes to data center cooling. ServerLSL replaces less efficient air coolers in the servers with redundant coolers (cold plate/pumps) and exhausts 100% of this hot air into the data center via heat exchangers (HEXs) in each server. This enables high wattage server nodes to have 1U form factors and maintain high cluster rack densities. At a site level, the heat is handled by existing CRACs and chillers with no changes to the infrastructure. With ServerLSL, liquid cooled nodes can be mixed in racks with traditional air-cooled nodes.

Asetek ServerLSL™ Cooling

While ServerLSL isolates the system within each server, Asetek RackCDU systems are rack-level focused, enabling a much greater impact on cooling costs of the datacenter overall. RackCDU systems leverage the same pumps and coolers used with ServerLSL nodes. RackCDU is in use by all of the current sites in the TOP500 using Asetek liquid cooling.

Asetek RackCDU provides the answer both at the node level and for the facility overall. As with ServerLSL, RackCDU D2C (Direct-to-Chip) utilizes redundant pumps/cold plates atop server CPUs & GPUs (and optionally other high wattage components like memory).  But the collected heat is move it via a sealed liquid path to heat exchangers in the RackCDU for transfer into facilities water. RackCDU D2C captures between 60% and 80% of server heat into liquid, reducing data center cooling costs by over 50% and allowing 2.5x-5x increases in data center server density.

The remaining heat in the data center air is removed by existing HVAC systems in this hybrid liquid/air approach. When there is unused cooling capacity available, data centers may choose to cool facilities water coming from the RackCDU with existing CRAC and cooling towers.

The high level of flexibility in addressing cooling at the server, rack, cluster and site levels provided by Asetek distributed pumping is lacking in approaches that utilize centralized pumping. Asetek’s approach continues to deliver flexibility in the areas of heat capture, coolant distribution and heat rejection.

At SC17, Asetek will also have on display a new cooling technology in which servers share a rack mounted HEX. The servers utilizing this shared HEX approach allow them to continue to be used if the site later moves to RackCDU.

To learn more about Asetek liquid cooling, stop by booth 1625 at SC17 in Denver.

Appointments for in-depth discussions about Asetek’s data center liquid cooling solutions at SC17 may be scheduled by sending an email to questions@asetek.com.


The post Trends in HPC and Machine Learning Drive Evolution of Cooling Solutions appeared first on HPCwire.

Arctur Offers Access to HPC infrastructure with HPC Challenge

Fri, 11/03/2017 - 09:42

Nov. 3, 2017 — ARCTUR has announced that €350,000 of ARCTUR HPC resources will be granted to the most interesting applications for the “Be innovative! HPC CHALLENGE” competition. The competition is open for applications between November 6th and December 1st 2017. ARCTUR will identify the winning applications by mid-December and the winning projects are expected to start in January of 2018.

The goal of the competition is to encourage small and medium enterprises (among others) to scale up their computer simulations and modeling by using ARCTUR HPC infrastructure, writing an interesting use case that demonstrates the benefits of HPC and becoming strategic partners of the company.

Applicants must submit a simulation or computer modeling idea addressing a real use case and demonstrating a high potential benefit from using ARCTUR HPC infrastructure.

The applications will be ranked based on these key selection criteria:

  • An application addressing a real use case
  • The HPC viability and scaling up of the case
  • Added benefits expected by using HPC
  • Applicants’ experience in using HPC (or GRID) computing infrastructures
  • Repeatability

Out of all the received applications, ARCTUR will select up to 30 cases that will receive access to Arctur-2 HPC infrastructure. Based on the success of the application, ARCTUR will provide one of the three different subsidy levels:

  1. Top 10 applications: Free access
  2. Second best 10 applications: 50% subsidy
  3. Third best 10 applications: 25% subsidy

Included HPC Services

Each application is expected to use up to €20,000 in resources in HPC and cloud capabilities or in hours of support. Usage will be limited to a maximum of 3 months of time.

ARCTUR supports the use of HPC within the framework of the Fortissimo, CloudFlow and CAxMan European projects. You can find the details of various cases on the links below:

The competition details are published here: https://arctur.si/hpc-and-cloud/hpc-challenge.

Arctur team photo


ARCTUR d.o.o. is a privately owned enterprise with its own HPC infrastructure located in Nova Gorica, Slovenia, Europe. Arctur’s On-Demand HPC gives true computing power and pay-as-you-go pricing. The compute environment includes a complete HPC software stack, a non-virtualized compute cluster, GPUs, high-memory notes, high-performance storage, and expert support. It is the leading service provider of supercomputing in Europe.

Časnik Finance and Manager Magazine recognized Arctur Silver as one of the Top 50 information technology companies in Slovenia. The company was awarded the IDC Innovation Excellence Award for Outstanding Application of HPC in 2014 (Pipistrel) and 2016 (Ergolines).

The company has extensive experience in server virtualization and deployment, integration of disparate IT systems, IT support of project management and server farm leverage. It is a pioneer in adapting HPC services to the needs of small and medium enterprises (SME).

Source: ARCTUR

The post Arctur Offers Access to HPC infrastructure with HPC Challenge appeared first on HPCwire.

Student Cluster Competition Plus Gambling? Count me in!

Fri, 11/03/2017 - 09:27

We all love HPC, right? We love college students too; particularly undergraduates that are striving to learn the ins and outs of HPC systems and applications – like in the SC17 Student Cluster Competition. It’s a fantastic combination.

But is it possible to make this event even more exciting? Yes, it is. Add gambling to the mix.

I’ve put together a SC17 Student Cluster Competition online betting pool. You get a virtual $1,000 to bet on any team you’d like, or even spread your bet among several teams. I’ll be updating the odds in future articles, just so you know how the field is shaping up. Bet wisely and have fun!!

Here’s the link to the betting pool.

The post Student Cluster Competition Plus Gambling? Count me in! appeared first on HPCwire.

SC17 Opening Plenary Announced: The Era of Smart Cities

Fri, 11/03/2017 - 07:58

DENVER, Nov. 3, 2017 — SC17 has announced that its opening plenary will be “The Era of Smart Cities — Reimagining Urban Life with HPC.”

When: Monday, November 13, 2017 @ 5:30 p.m. MT, Denver, Colorado Convention Center

Access: Open to anyone with a SC17 conference badge

What: The 21st Century is frequently referenced as the “Century of the City,” reflecting the unprecedented global migration into urban areas. Urban areas are driving and being affected by factors ranging from economics to health to energy to climate change. The notion of a “smart” city is one that recognizes the use and influence of technology in cities. It is the wave of the future only made possible by high performance computing (HPC).

This expert plenary panel will discuss emerging needs and opportunities suggesting an increasing role for HPC in cities, with perspectives from city government, planning and design, and embedded urban HPC systems.

Why: HPC is already playing a key role in helping cities pursue objectives of societal safety, efficient use of resources, and an overall better quality of life. Intelligent devices enabled with HPC “at the edge” have potential to optimize energy generation and delivery, emergency response or the flow of goods and services, and to allow urban infrastructures to adapt autonomously to changes and events such as severe storms or traffic congestion. Smart city technology can even improve food safety inspections and help identify children most at risk for lead poisoning.

HPC is supporting the creation of breakthrough computational models to make all of this possible. Hear from industry experts who are pioneers in making these life-changing realties happen today.

Moderator: Charlie Catlett, Director, Urban Center for Computation & Data, Argonne National Laboratory

Panelists: Debra Lam, Managing Director for Smart Cities & Inclusive Innovation, Georgia Tech; Michael Mattmiller, Chief Technology Officer, the City of Seattle; Pete Beckman, Director, Co-Director, Northwestern Argonne Institute of Science and Engineering, Argonne National Laboratory

Source: SC17

The post SC17 Opening Plenary Announced: The Era of Smart Cities appeared first on HPCwire.

SC17 Cluster Competition Teams Unveiled!

Thu, 11/02/2017 - 16:54

As we draw nearer to SC17, tensions are rising with each passing day. Sixteen teams of university undergrads are putting the final touches on their clusters and preparing for the grueling test that is the SC17 Student Cluster Competition.

Nothing comes easy in the student clustering game, and each team will have to be at the top of their game in order to have a chance for the SC17 Student Cluster Crown (although there’s not an actual crown.)

Millions of fans will be tracking the competition as the teams vie to turn in the best system performance on a suite of HPC benchmarks and applications – all while staying under the 3,000 watt power cap. For those of you who might not be familiar with how these competitions work, check out our overview here.

I recorded a webcast with Stephen Harrell, the chair of the cluster competition this year and an all-around Student Cluster Competition guru. In the video, we introduce the teams that will be competing this year and discuss their various approaches to the competition.


We’re going to be all over the competition this year, covering it through articles, videos, and perhaps even an interpretive dance or off-Broadway musical. Stay tuned for more.

The post SC17 Cluster Competition Teams Unveiled! appeared first on HPCwire.

SCinet at SC17: Meet the World’s Fastest and Most Powerful Temporary Network

Thu, 11/02/2017 - 16:22

Nov. 2 — Most high-speed, high-bandwidth networks consist of fiber optic cable that is permanently installed in protective walls, ceilings and flooring, or buried underground in thick, waterproof conduit. But building the world’s fastest and most powerful  temporary network from the ground up each year for the SC conference necessitates a different approach.

In the days leading up to SC17, more than 180 volunteers will gather in Denver to set up SCinet, the high-capacity network that supports the revolutionary applications and experiments that are a hallmark of the SC conference. SCinet takes one year to plan, and those efforts culminate in a month-long period of staging, setup and operation of the network during the conference.

SCinet volunteers temporarily installed more than 100 miles of optical fiber in the Salt Palace Convention Center in less than one week in preparation for SC16 in Salt Lake City. Carpet and masonite hardboard were placed over the fiber to protect it from heavy equipment and the foot traffic of more than 11,000 attendees and exhibitors.

Miles of fiber will be temporarily taped to the cement floors and hung from the rafters of the Colorado Convention Center. Although carpet and rigid masonite will cover the fiber on the ground, it first will be exposed to the traffic of dozens of exhibitor carts, forklifts and other utility vehicles that pass nearby and sometimes over it via temporary thresholds as exhibits are constructed on the show floor.

Rebecca Hutchinson and Annette Kitajima are SCinet volunteers who co-chair the fiber team. Hutchinson said the need for on-site fiber repairs on the ground at SC varies greatly from year to year.

“In my eight years on the fiber team, we’ve had at least one conference where no repairs were required from setup to teardown,” Hutchinson said. “Another year, we were repairing fiber in the middle of the show floor as the exhibits opened on the second day.”

The fiber team uses fusion splicing (using heat to fuse the ends of two optical fibers), section replacement and hand termination to repair connections as quickly as possible, so disruptions are minimal.

When the conference closes on November 17, all of that fiber will be removed in less than 24 hours.

“Even if the fiber makes it through an SC unscathed, re-spooling at the end of the show takes its toll, with the inherent twists and kinks that come with a speedy teardown,” Kitajima said.

Members of the team will test all of the fiber during SCinet’s annual inventory meeting in the spring. Some of it will be damaged beyond repair, but most of it will be redeployed next year at SC18 in Dallas, just as most of the fiber in Denver is being reused from SC16 in Salt Lake City. This is part of SCinet’s sustainability strategy – recovering, refurbishing and reusing available resources whenever possible.

Brad Miller and Jerry Beck of the Utah Education and Telehealth Network are fiber experts and SCinet volunteers who spent additional time during the summer and fall testing and repairing a batch of used fiber from past SCs. Miller said that of the 84 reels the duo has tested, 24 have been refurbished. That amounts to 28 percent or 1.28 miles.

Beck starts the refurbishing process with an EXFO OTDR, or Optical Time Domain Reflectometer, to determine how well a given reel of used fiber will transmit light. The reflectometer measures the speed of light as it passes through the fiber to identify flaws and pinpoint “pretty much exactly where problems will be.” In most cases, replacing the ends of the cable resolves the issue.

To attach new connectors he uses a Swift F1 cable splicer. It expertly strips protective insulation and exposes fiber thinner than human hair, and then aligns, fuses and reinsulates it.

“Sometimes you can be doing splicing with this machine and a hair will fall into the splice area and it will indicate that it’s a fiber and will say your fiber isn’t clean,” Beck said. “All these fibers are measured in microns. It takes the actual fiber cores and lines them up so close that it tells me that the clad is three microns or three millionths of an inch.”  Beck said the work does involve an element of danger: Tiny unshielded fiber fragments can easily pierce fingers and are particularly painful if accidentally rubbed into an eye.

Watch Brad Miller and Jerry Beck demonstrate the fiber refurbishing process in this 3:44 video: https://youtu.be/FEls2MKUIBY or just watch below:

Source: Brian Ban, SC

The post SCinet at SC17: Meet the World’s Fastest and Most Powerful Temporary Network appeared first on HPCwire.

ACSCI at SC17: Students Explore Immigration Through a Big Data Lens

Thu, 11/02/2017 - 15:36

Nov. 2 — Supercomputers have helped scientists discover merging black holes and design new nanomaterials, but can they help solve society’s most challenging policy issues?

At the International Conference for High Performance Computing, Networking, Storage and Analysis (also known as Supercomputing 2017 or SC17) in Denver, Colorado, from Nov. 12 to Nov. 15, undergraduate and graduate students from diverse disciplines and backgrounds will learn how to use advanced computing skills to explore the nation’s immigration policies as part of the Advanced Computing for Social Change Institute (ACSCI).

Daring Greatly, SC16 Advanced Computing for Social Change facilitators and participants. (Photo by Marques Bland)

The program, which debuted in 2016, teaches students how to use data analysis and visualization to identify and communicate data-driven policy solutions to pressing social issues. Organized by the Texas Advanced Computing Center (TACC), the Extreme Science and Engineering Discovery Environment (XSEDE, a National Science Foundation-funded organization), and SC17 the project is unique in its application of the world’s most cutting-edge technologies to address social change.

“The institute will help students realize their leadership potential and increase their confidence in their ability to effect social change, regardless of where their profession takes them in the future,” said Rosalia Gomez, TACC education and outreach manager and one of the organizers of the program.

“Our goal is to provide students with advanced computing skills and the ability to visualize complex data in a way that is useful to everyone, from those affected by immigration policy to those creating immigration policy,” said Ruby Mendenhall, an associate professor of sociology at the University of Illinois and another of the organizers. “It is our hope that students will provide new knowledge about U.S. immigration that can create social change.”

During the three-and-a-half-day program, students will learn computing, data analysis and visualization skills, and then form teams to tackle critical issues related to immigration. The students take a data-driven approach, analyzing large datasets to derive persuasive arguments for their positions.

“These students have so much to offer to the national conversation and debate,” said Kelly Gaither, director of visualization at TACC. “Data analysis and visualization provide a vehicle to get to the truth behind the rhetoric. While this is an emotionally charged topic, we can use advanced computing tools to discern fact from fiction and use this as a platform for constructive communication to move us forward.”

As part of the projects students will also receive mentorship, career development advice, and network with other advanced computing professionals. At the end of the program, the teams will present their analyses to experts in the supercomputing field.

Last year, at SC16 in Salt Lake City, Utah, students explored the Black Lives Matter movement and developed arguments advocating greater national unity and an end to police violence against people of color. The majority of student participants were from under-served communities and were first generation college students.

“The Advanced Computing for Social Change Institute aims to empower young minds in utilizing unique software and their talents for modeling and analyzing everyday social issues irrespective of their academic major, thus effectively making predictions pertaining to the inspired model,” said Stacyann Nelson, a graduate student at Florida A&M University and a part participant and current mentor in the program.

Participants in this year’s event come from colleges and universities across Colorado, including the University of Colorado Boulder, the University of Denver, the Colorado School of Mines, and the University of Colorado Denver.

“I am thrilled that several computer science students from the Colorado School of Mines will be participating in this important event,” said Tracy Camp, professor of Computer Science at the university. “These types of events are a win-win, as they offer an opportunity for our students to increase their skills and an opportunity for their efforts to impact the world in a positive way. The Colorado School of Mines is grateful that the event is just down the road in Denver.”

Link to original article with additional images.

Source: TACC

The post ACSCI at SC17: Students Explore Immigration Through a Big Data Lens appeared first on HPCwire.

Nvidia Charts AI Inroads, Inference Challenge at DC Event

Thu, 11/02/2017 - 14:36

As developers flock to artificial intelligence frameworks in response to the explosion of intelligence machines, training deep learning models has emerged as a priority along with syncing them to a growing list of neural and other network designs.

All are being aligned to confront some of the next big AI challenges, including training deep learning models to make inferences from the fire hose of unstructured data.

These and other AI developer challenges were highlighted during this week’s Nvidia GPU technology conference in Washington. The GPU leader uses the events to bolster its contention that GPUs—some with more than 5,000 CUDA cores—are filling the computing gap created by the decline of Moore’s Law. The other driving force behind the “era of AI” is the emergence of algorithm-driven deep learning that is forcing developers to move beyond mere coding to apply AI to a growing range of automated processes and predictive analytics.

Nvidia executives demonstrated a range of emerging applications designed to show how the parallelism and performance gains of GPU technology along with the company’s CUDA programming interface complement deep learning development. Nvidia said Wednesday (Nov. 1) downloads of the CUDA API have doubled over the past year to 2 million.

“Becoming the world’s AI [developer] platform is our focus,” asserted Greg Estes, Nvidia’s vice-president for developer programs. Buttressing that strategy, the company also this week announced expansion of its Deep Learning Institute to address the growing need for more AI developers.

As it seeks to put most of the pieces in place to accelerate AI development, Nvidia has also pushed GPU technology into the cloud with early partners such as Amazon Web Services. The company’s GPU cloud seeks to create a “registry” of tools for deep learning development. Estes added that the company has taken a hybrid cloud approach in which the deep learning registry runs in the cloud or on premises while deep learning frameworks such as Tensor Flow can be delivered via application containers.

Last week Amazon became the first major public cloud provider to offer the top-shelf Nvidia Tesla Volta GPUs.

As the deep learning ecosystem comes together, Estes argued in a keynote address that the training of models to infer from huge data sets loom large. “AI inference is the next great challenge,” he argued. “It turns out the problem is pretty hard.” The scale of the challenge was illustrated by this statistic: There are an estimated 20 million inference servers currently crunching data to make inferences that run the gamut from educated guesses to reliable predictions.

Estes ticked off a growing list of emerging network designs that underpin current deep learning development, ranging from convolutional networks for visual data and recurrent nets for speech recognition to reinforcement learning and generative adversarial networks (in which two opposing networks seek to “fool” the other to, for example, spot a forgery).

Hence, Nvidia and its growing list of cloud and developer partners have been laboring to apply GPU parallelism and deep learning frameworks like TensorFlow to accelerate model training. The company released the latest version of its TensorRT AI inference software last month. The combination of Tesla GPUs and CUDA programmability are designed to “accelerate the growing diversity and complexity of deep neural networks,” Nvidia CEO Jensen Huang asserted.

Nvidia also uses its roadshows to demonstrate its GPU capabilities versus current CPUs. An inferencing example was intended to back its GPU performance claims. The demonstration involved sorting through aerial photos that were labeled according to land uses such as agriculture of an airstrip.

The company claims its approach could map an area the size of New Mexico in a day while a CPU platform would require three months to sort and organize the aerial photos.

Emerging deep learning tools could be applied to applications such as remote sensing and other satellites imagery where emerging vendors are struggling to sift through hundreds of petabytes of data.

Nvidia partner Orbital Insights Inc. said it is combining satellite imagery with deep learning tools for applications like determining crop yields and urban planning. CEO James Crawford said it is using GPUs and deep learning tools such as convolutional models to process images taken by synthetic aperture radars that are able to see through clouds.

As part of an effort to quantify Chinese energy output, the San Francisco-based startup trained a neural network to spot the natural gas flares associated with fracking. Among the unintended but valuable finds delivered by the deep learning model was uncovering details of China’s industrial capacity, including the number of blast furnaces used in steel production.

The post Nvidia Charts AI Inroads, Inference Challenge at DC Event appeared first on HPCwire.

ECP Pagoda Project Rolls Out First Software Libraries

Thu, 11/02/2017 - 13:48

Nov. 2 — Just one year after the U.S. Department of Energy’s (DOE) Exascale Computing Program (ECP) began funding projects to prepare scientific applications for exascale supercomputers, the Pagoda Project — a three-year ECP software development program based at Lawrence Berkeley National Laboratory — has successfully reached a major milestone: making its open source software libraries publicly available as of September 30, 2017.

Led by Scott B. Baden, Group Lead of the Computer Languages and Systems Software (CLaSS) Group within Berkeley Lab’s Computational Research Division, the Pagoda Project’s libraries are designed to support lightweight global address space communication for exascale applications. The libraries take advantage of the Partitioned Global Address Space (PGAS) model to emulate large, distributed shared memories. By employing this model, which allows researchers to treat the physically separate memories of each supercomputer node as one address space, the Pagoda libraries will be able to leverage available global address hardware support to significantly reduce the communication costs of moving data — often a performance bottleneck in large-scale scientific applications, Baden explained.

“Our job is to ensure that the exascale applications reach key performance parameters defined by the DOE,” he added.

Thus this first release of the software is as functionally complete as possible, Baden emphasized, covering a good deal of the specification released last June. “We need to quickly determine if our users, in particular our ECP application developer partners, are satisfied,” he said. “If they can give us early feedback, we can avoid surprises later on.”

GASNet-EX and UPC++

The Pagoda software stack comprises a communication substrate layer, GASNet-Ex, and a productivity layer, UPC++. GASNet-Ex is a communication interface that provides language-independent, low-level networking for PGAS languages such as UPC and Coarray Fortran, the UPC++ library and for the Legion Programming Language. UPC++ is a C++ interface for application programmers that creates “friendlier” PGAS abstractions above GASNet-Ex’s communication services.

“GASNet-Ex, which has been around for over 15 years and is being enhanced to make it more versatile and performant in the exascale environment, is a library intended for developers of tools that are in turn used to develop applications,” Baden explained. “It operates at the network hardware level, which is more challenging to program than at the productivity layer.” The GASNet-Ex effort is led by Pagoda co-PI Paul Hargrove and was originally designed by Dan Bonachea, who jointly develops the software. Both are members of CLaSS.

As the productivity layer, UPC++ sits at a slightly higher level over GASNet-Ex, in a form appropriate for applications programmers. The goal of this layer is to impose minimal overheads in exchange for hiding considerable idiosyncratic detail, so users are satisfied with the benefits obtained by increased productivity.

Over the past year, the Pagoda team worked closely with several Berkeley Lab partners to develop applications and application frameworks, including the Adaptive Mesh Refinement Co-Design Center (AMReX), Sparse Solvers (ECP AD project) and ExaBiome (ECP AD Project). They also worked with several industry partners, including IBM, NVIDIA, HPE and Cray, and over the next few months will be meeting with all of the major vendors who are vying to build the first exascale computer or the components that will go into those computers.

“We are part of a large community of ECP developers,” Baden said. “And the ECP wants to deploy a software stack, a full set of tools, as an integrated package that will enable them to ensure that the pieces are compatible, that they will all work together. I am fortunate to be working with such a talented team that is highly motivated to deliver a vital component of the ECP software stack.” This team includes other members of CLaSS—Steve Hofmeyr and Amir Kamil (at the University of Michigan)—as well John Bachan, Brian van Straalen and Mathias Jacquelin. Bryce Lelbach, now with NVIDIA, also made early contributions.

Now that they are publicly available, the Pagoda libraries are expected to be used by other ECP efforts and supercomputer users in general to meet the challenges posed not only by the first-generation exascale computers but by today’s petascale systems as well.

“Much of the ECP software and programming technology can be leveraged across multiple applications, both within ECP and beyond,” said Kathy Yelick, Associate Lab Director for Computing Sciences at Berkeley Lab, in a recent interview with HPCwire. For example, AMReX, which was launched last November and recently announced its own first milestone, released its new framework to support the development of block-structured AMR algorithms, and at least five of the ECP application projects are using AMR to efficiently simulate fine-resolution features, Yelick noted.

For the remaining two years of the Pagoda project, the team will be focused on application integration and performance enhancements that adeptly leverage low-level hardware support, Baden noted.

Source: Berkeley Lab

The post ECP Pagoda Project Rolls Out First Software Libraries appeared first on HPCwire.

UberCloud Brings Parallel MPI to Univa Grid Engine’s Docker Support

Thu, 11/02/2017 - 13:41

CHICAGO, Nov. 2 2017 – UberCloud and Univa today announced the integration of UberCloud parallel application containers with Univa Grid Engine. In May of last year, Univa, a leading innovator of workload management products, announced the availability of Docker software container support with its Grid Engine 8.4.0 product, enabling enterprises to automatically dispatch and run jobs in Docker containers, from a user specified Docker image, on a Univa Grid Engine cluster.

This significant update simplifies running complex applications in a Univa supported cluster and reduces configuration and OS issues. Since then, user applications are isolated into their own container, avoiding conflicts with other jobs on the system. This integration enables Docker containers and non-container applications to run in the same cluster, on-premise or in the cloud.

UberCloud and Univa technologies complement each other with Univa focusing on orchestrating jobs and containers on any computing resource infrastructure. Univa and UberCloud have worked together to bring MPI parallel execution to Univa Grid Engine managed containers. Parallel applications are pre-installed in UberCloud containers which are managed by Univa in the same way jobs are managed.

Now Univa and Unicloud solutions can automatically launch containerized parallel workloads in the cloud when the local system queues are full. UberCloud containers guarantee that the same software version can run on-premise and in the cloud. Also, software upgrades become automated, and the users just specify which software version of e.g. ANSYS Fluent or SIMULIA Abaqus they want, by selecting the container with the software version they need. Finally, for interactive sessions, the UberCloud/Univa integration provides Nice-DCV and VNC support for GPU-accelerated remote graphics.

Univa’s suite includes the world’s most trusted workload optimization solution enabling organizations to manage and optimize distributed applications, containers, data center services, legacy applications, and Big Data frameworks in a single, dynamically shared set of resources. It is the most widely deployed workload optimization solution used today in more than 10,000 data centers across thousands of applications and use cases.

Be sure to stop by Univa booth (#849) at SC17 to learn more about containers on November 13 – 16, in Denver, Colo.

About UberCloud

UberCloud is the online Community, Marketplace, and Software Container Factory where engineers, scientists, and their service providers, discover, try, and buy ubiquitous high-performance computing power and Software-as-a-Service, from Cloud resource providers and application software vendors around the world. UberCloud’s unique high-performance software container technology simplifies software packageability and portability, enables ease of access and instant use of engineering SaaS solutions, and maintains scalability across multiple compute nodes. Please visit www.TheUberCloud.com or contact us at www.TheUberCloud.com/help/.

About Univa Corporation

Univa is the leading innovator of workload management products that optimize performance of applications, services and containers. Univa enables enterprises to fully utilize and scale compute resources across on-premise, cloud, and hybrid infrastructures. Advanced reporting and monitoring capabilities provide insights to make scheduling decisions and achieve even faster time-to-results. Univa’s solutions help hundreds of companies to manage thousands of applications and run billions of tasks every day. Univa is headquartered in Chicago, with offices in Canada and Germany. For more information, please visit www.univa.com.

Source: Univa Corp.

The post UberCloud Brings Parallel MPI to Univa Grid Engine’s Docker Support appeared first on HPCwire.

LiquidCool Solutions Names Darwin P. Kauffman CEO

Thu, 11/02/2017 - 10:47

ROCHESTER, Minn., Nov. 2, 2017 — LiquidCool Solutions today announced that Darwin P. Kauffman has been appointed as the company’s new CEO, effective immediately. The announcement was made today by former CEO Herb Zien and Chairman Stephen Einhorn, both of whom will remain as co-chairmen of the Board.

“We are delighted that Darwin will lead LiquidCool into its next stage of growth,” said Zien. “With Darwin’s stellar technical, product and leadership credentials, he will be instrumental in the development and execution of LiquidCool’s long-term success.”

Kauffman, who has a degree in electrical engineering and an MBA, began his career at Seagate Technology where he led product development, product management and strategy roles for Seagate’s enterprise storage devices and solutions products. While with Seagate, Kauffman transformed one of its declining $700M+ business unit into a $1B+ enterprise; launched enterprise HDD products into existing market capturing 12% market share and $800M in new revenue; and consistently drove double and triple-digit revenue and profit growth in various technology market segments.

“This is an exciting time for LiquidCool Solutions, and I am thrilled to be taking on the CEO role,” said Kauffman. “LiquidCool solidified the total immersion server industry with the most elegant and efficient liquid cooling technology for electronic equipment. I’m honored and excited about leading LiquidCool into our next phase of growth, where we will combine business and product innovation to provide even more customer value.”

About LiquidCool Solutions

LiquidCool Solutions is a technology development firm with 30 issued patents centering on cooling electronics by total immersion in a dielectric liquid. LCS technology places special emphasis on scalability and rack management.  Beyond providing superior energy savings, performance and reliability, LCS technology enables a broad range of unique applications not possible with any other air or liquid cooling systems.

Source: LiquidCool Solutions

The post LiquidCool Solutions Names Darwin P. Kauffman CEO appeared first on HPCwire.

SC17 Preview: The National Strategic Computing Initiative

Thu, 11/02/2017 - 10:02

In Washington, the conventional wisdom is that an initiative started by one presidential administration will not survive into a new one. This seemed to be particularly true with the transition of the Obama administration into the Trump administration. However, an exception to this unwritten rule may be the case of an initiative to support exascale, data analytics, “post-Moore’s Law” computing and the HPC ecosystem. The jury is still out, but the signs are starting to look good.

In the summer of 2014, during the tail-end of the Obama administration, a team at the White House’s Office of Science and Technology Policy (OSTP) started to formulate what would become known as the National Strategic Computing Initiative (NSCI). Over the next year, the NSCI was defined and refined through an interagency process and interactions with computer companies and industry users of high performance computing. Although the initiative was formally started by President Obama on July 29, 2015, support by the US federal government for advanced computing is not new, nor is the concept of multi-agency national strategic computing programs. For example, precedents include the Strategic Computing Initiative of the 1980s, the High-Performance Computing Act of 1991, and the High Productivity Computing Systems program of the 2000s. Information concerning NSCI can be found at https://www.nitrd.gov/nsci/.

NSCI recognizes the value of US investments in cutting-edge, high-performance computing for national security, economic security, and scientific discovery. It directs the administration to take an “whole of government” approach to continuing and expanding those activities. The initiative puts the Department of Defense, the Department of Energy and the National Science Foundation into leadership roles to coordinate those efforts. The initiative also identifies different agencies to conduct foundational R&D and be involved with deployment and implementation. The “whole of government” approach is quite important to collect and coordinate the resources (i.e. funding) to achieve the NSCI goals.

There are five strategic objectives for this initiative. The first is to accelerate the delivery of a “capable exascale computing system” (defined as the integration of hardware and software capability to deliver approximately 100 times the performance of current 10-petaflop systems across a range of applications representing government needs). The second seeks to increase the coherence between traditional modeling and simulation and large data analytics. The third objective is to establish, over the next 15 years, a viable path forward for advanced computing in the “post Moore’s Law era.” The fourth objective seeks to increase the capacity and capability of the entire HPC ecosystem, both human and technical. Finally, the fifth NSCI objective is to implement enduring public-private collaborations to ensure that the benefits of the initiative are shared between the government and the industrial and academic sectors of the economy.

An NSCI Joint Program Office (JPO) has been established with representatives from the lead agencies (DOD, DOE, and NSF). There was also a decision to have the Networking and Information Technology Research and Development (NITRD)’s National Coordination Office (NCO) to act as the communications arm for the initiative. Also, an Executive Council led by the directors of OSTP and the OMB (Office of Management and Budget) has been established and in July of 2016 published a Strategic Plan for the initiative.

The bad news is that there were not any formally designated funds for NSCI identified in the President Trump’s Fiscal Year 2018 request (although the initiative was mentioned in several places). In the federal government that could be the “kiss of death.” An initiative without funding often withers away and dies. The encouraging thing about the NSCI is that it may be okay that there is no specifically designated funding. The reason for this is that there other currently funded activities at the lead agencies that already align with the goals of the NSCI. Therefore, the only thing needed for “NSCI implementation” is for these activities to work in a coordinated way and that is already happening, to some degree, through the JPO. The synergy of the currently funded NSCI relevant activities provides additional hope that the initiative will survive the transition.

Other pieces of good news include the fact that the staff at the White House’s OSTP is growing and we understand has been briefed on the initiative. We also heard that the White House’s Deputy Chief Technology Officer, Michael Kratsios, has been briefed on NSCI. Another very good sign was that on August 17th, Mike Mulvaney of OMB and Michael Kratsios issued the Administration’s R&D budget priorities. One of those, under the category of Military Superiority, was the call for the U.S. to maintain its leadership in future computing capabilities. Also, under the category of American Prosperity, the budget priorities expressed an interest in R&D in machine learning and quantum computing. Finally, there was direction given for the coordination of new R&D efforts to avoid duplication with existing efforts, which is what the NSCI JPO is already doing.

More specific information about the status of the NSCI will be available at the upcoming Birds of a Feather session at the SC17 conference (5:15 pm, Wed 11/15, Room 601). There, current members of the JPO (Mark Sims of DOD, William Harrod of DOE, and Irene Qualters of NSF) will be able to provide the latest and greatest about the initiative.

For the initiative to survive, the new administration will need to take ownership. Sometimes, with an administration shift, this may involve adjusting its scope. However, there has been previous initiatives that successfully made the administration leap intact (an example is the DOE Accelerated Strategic Computing Initiative (ASCI)). These tend to be initiatives that have a clear and compelling reason to exist and a sound organization that provides confidence that they will succeed.

Things continue to look good for funding the exascale program in the Trump administration. Also, the growth of large scale data analytics across the spectrum of government, industry, and academia probably means that there is a good chance that NSCI will survive the transition.

About the Author

Alex Larzelere is a senior fellow at the U.S. Council on Competitiveness, the president of Larzelere & Associates Consulting and HPCwire’s policy editor. He is currently a technologist, speaker and author on a number of disruptive technologies that include: advanced modeling and simulation; high performance computing; artificial intelligence; the Internet of Things; and additive manufacturing. Alex’s career has included time in federal service (working closely with DOE national labs), private industry, and as founder of a small business. Throughout that time, he led programs that implemented the use of cutting edge advanced computing technologies to enable high resolution, multi-physics simulations of complex physical systems. Alex is the author of “Delivering Insight: The History of the Accelerated Strategic Computing Initiative (ASCI).”

The post SC17 Preview: The National Strategic Computing Initiative appeared first on HPCwire.

Appentra Raises a Funding Round of €400K

Thu, 11/02/2017 - 09:15

Nov. 2, 2017 — Appentra, a technology-based company spin-off of the University of Coruña established in 2012, has raised a funding round of €400.000. The round was led by three Spanish venture capital organizations: Caixa Capital Risc, Unirisco and Xesgalicia.

The new funds will be used to accelerate the market uptake of Parallware tools, allowing us to scale our team and to further improve the Parallware technology.

Appentra  provides top quality software tools that allow extensive use of High Performance Computing (HPC) techniques in all application areas of engineering, science and industry. Appentra’s target clients are companies and organizations that run frequently updated compute-intensive applications in markets like aerospace, automotive, civil engineering, biomedicine or chemistry.

”It is a privilege to have such a supportive group of investors that believe in our vision and in our team.” said Manuel Arenaz (CEO).

About Caixa Capital Risc

Caixa Capital Risc is the venture capital arm of Criteria Caixa, an investor that provides equity and convertible loans to innovative companies in their early stages. They manages a capital of 195 million euro and invests mainly in Spanish companies in the Industrial technology, Healthcare/Life Sciences and time fields.

About Unirisco

Unirisco is a venture capital group promoting the creation of companies making use of university knowledge. This is achieved through short-term investment operations in their financing or through other financial instruments, always with the criteria of profitability and job creation in mind.

About Xesgalicia

Xesgalicia is a Galician Venture Capital Management firm. It finances company development through the temporary acquisition of minority shares of the capital of unquoted companies. In addition, it may make ordinary or mezzanine loans to the companies in which it invests through differents venture capital funds and the assets of a venture capital company.

Source: Appentra

The post Appentra Raises a Funding Round of €400K appeared first on HPCwire.

NSF Selects Anne Kinney to Head Mathematical and Physical Sciences Directorate

Thu, 11/02/2017 - 09:08

Nov. 2, 2017 — The National Science Foundation (NSF) has selected Dr. Anne Kinney to serve as head of the Directorate for Mathematical and Physical Sciences (MPS), which supports fundamental research in astronomy, chemistry, physics, materials science and mathematics.

Kinney has more than 30 years of leadership and management experience in the astronomical community. Since 2015, she has been serving as chief scientist at the W. M. Keck Observatory, which hosts the world’s largest optical and infrared telescopes. At Keck, she served as a liaison to the global scientific community, acting as an ambassador to the observatory’s entire user community.

Prior to that, Kinney held multiple positions at NASA’s Goddard Space Flight Center — most recently as Director of the Solar System Exploration Division, leading and managing a team of more than 350 people. Before moving to Goddard Space Flight Center, Kinney was director of the Universe Division at NASA Headquarters. She oversaw successful space missions that included the Hubble Space Telescope, the Spitzer Space Telescope, the Wilkinson Microwave Anisotropy Probe and the Galaxy Evolution Explorer.

“Anne Kinney arrives at a special moment in our quest to understand the universe — as excitement builds for a new era of multi-messenger astrophysics. And, as we look to convergence research to address some of the most challenging issues in science and engineering, all of the fields in the MPS directorate — mathematics, chemistry, materials science, physics and astronomy — play foundational and leading roles,” said NSF Director France Córdova. “Kinney has successfully brought together researchers, educators, students and other partners time and again to support significant scientific and engineering feats. I am thrilled to welcome her to the NSF leadership team, where her skills and experience will help us maintain our position keeping the U.S. at the forefront of scientific and technological excellence.”

MPS provides about 43 percent of the federal funding for basic research at academic institutions in the mathematical and physical sciences. The directorate serves the nation by supporting fundamental discoveries at the leading edge of science, with special emphasis on supporting early career investigators and advancing areas of science, including quantum information science, optics, photonics, clean energy, data science and more. The NSF-funded Laser Interferometer Gravitational-Wave Observatory (LIGO), which recently detected the collision of two neutron stars, has been supported by MPS for more than 40 years.

“MPS explores some of our most compelling scientific questions, and I am eager to add to the efforts of an agency that plays a key role in driving the U.S. economy, ensuring national security and enhancing the nation’s global leadership in innovation,” Kinney said. “Throughout my career, I’ve been fortunate to lead teams that have used knowledge gained from breakthroughs in fundamental science to enrich how we see and understand the universe. It’s exciting to think that my work at MPS will support research with the potential to fuel decades’ worth of future exploration.”

An expert in extragalactic astronomy, Kinney has published more than 80 papers on quasars, blazars, active galaxies and normal galaxies, and signatures of accretion disks in active galaxies. Her research showed that accretion disks in the center of active galaxies lie at random angles relative to their host galaxies.

Kinney has won numerous awards and honors, including the Presidential Rank Award for Meritorious Service, the NASA Medal for Outstanding Leadership and several NASA Group Achievement Awards for projects such as the Keck Observatory Archive and the James Webb Space Telescope, the Gamma-ray Large Area Space Telescope (now known as the Fermi Gamma-ray Space Telescope) and Lunar Orbiter Laser Altimeter. An avid supporter of science communication and outreach, Kinney created the Space Telescope Science Institute education group — launching the online Amazing Space program — and has served on the editorial board of Astronomy Magazine since 1997.

Kinney has a bachelor’s degree in astronomy and physics from the University of Wisconsin and a doctorate in astrophysics from New York University. She studied in Denmark for several years at the Niels Bohr Institute, University of Copenhagen.

Kinney will begin her NSF appointment on Jan. 2, 2018.

Source: NSF

The post NSF Selects Anne Kinney to Head Mathematical and Physical Sciences Directorate appeared first on HPCwire.

XSEDE Hosting Collaboration Booth at Supercomputing 17

Thu, 11/02/2017 - 09:06

Nov. 2, 2017 — The Extreme Science and Engineering Development Environment (XSEDE) will be hosting a collaboration booth at Supercomputing Conference (SC) in Denver, Colorado from November 13th through 16th.

XSEDE, which is supported by the National Science Foundation (NSF), is the most advanced, powerful and robust collection of integrated digital resources and services in the world. XSEDE is inviting SC17 attendees to the XSEDE collaboration booth (#225) to learn about XSEDE-enabled projects, chat with XSEDE representatives, and see what type of resources and collaborations XSEDE offers researchers. XSEDE’s collaboration booth also will feature two-half day events to encourage XSEDE meetups and project collaborations.

RSVP here:

Source: XSEDE

The post XSEDE Hosting Collaboration Booth at Supercomputing 17 appeared first on HPCwire.

Ellexus Launches Container Checker on AWS Marketplace

Thu, 11/02/2017 - 09:01

CAMBRIDGE, England, Nov. 2, 2017 — Ellexus has launched Container Checker on Amazon Web Services’ Marketplace, a pioneering cloud-based tool that provides visibility into the inner workings of Docker containers.

Container Checker is the first tool aimed directly at the cloud market from Ellexus, the I/O profiling company. The software provider has seven years’ experience in providing dependency analysis and performance tools in big-compute environments. Ellexus Container Checker brings this expertise to a much wider audience and will enable organisations of all sizes in many sectors to scale rapidly and quickly.

Using the tool is simple; spin up the container on an AWS machine, run the Container Checker trace and receive your report when the trace is complete. It will only take as long as the application takes to run.

The following checks are included in the trace:

  • I/O performance: Small reads and writes? Excess meta data operations? Discover all the flaws that are bringing your application to a standstill – and costing you in wasted cloud spend
  • Dependencies and security: What files does my application need to run? Are they inside the container or outside? Make sure you have everything you need and the container doesn’t access files or network locations that it shouldn’t
  • Time: Where is your application wasting time? Should you use a different cloud set-up? Find out if your container is a time waster.

Currently available on AWS, the tool will soon be rolled out across other cloud platforms. Over time Container Checker will also add more types of container to its checklist.

“We are extremely excited to launch our first tool aimed directly at improving the lives of cloud platform users,” said Ellexus CEO Dr Rosemary Francis.

“The use of both cloud compute clusters and containers is set to grow extremely quickly over the next few years, which undoubtedly means people will run into challenges as they adapt to new set-ups and try to scale quickly.

“Container Checker will help people using cloud platforms to quickly detect problems within their containers before they are let loose on the cloud to potentially waste time and compute spend. Estimates suggest that up to 45% of cloud spend is wasted due in part to unknown application activity and unsuitable storage decisions, which is what we want to help businesses tackle.”

Find out more at www.ellexus.com.

 About Ellexus Ltd

Ellexus is an I/O profiling company, trusted by the world’s largest software vendors and high performance computing organisations. Its monitoring and profiling tools profile thousands of applications daily, improving performance by up to 35%.

Source: Ellexus

The post Ellexus Launches Container Checker on AWS Marketplace appeared first on HPCwire.

Asetek Announces OEM Partnership With E4 Computer Engineering and Installation

Thu, 11/02/2017 - 08:26

OSLO, Norway, Nov. 2, 2017 — Asetek (ASETEK.OL) today announced E4 Computer Engineering, an Italian technology provider of solutions for HPC, data analytics and AI, as a new data center OEM partner. E4 Computer Engineering has utilized Asetek RackCDU D2C (Direct-to-Chip) liquid cooling for the D.A.V.I.D.E. SUPERCOMPUTER in Italy. This follows the announcement of an undisclosed OEM partner and installation on July 14, 2017.

You can read more about the partnership at Asetek.com.

About Asetek

Asetek is a global leader in liquid cooling solutions for data centers, servers and PCs. Founded in 2000, Asetek is headquartered in Denmark and has operations in California, Texas, China and Taiwan. Asetek is listed on the Oslo Stock Exchange (ASETEK). For more information, visit www.asetek.com

Source: Asetek

The post Asetek Announces OEM Partnership With E4 Computer Engineering and Installation appeared first on HPCwire.

Huawei Partners with Intel to Build a Supercomputing Cluster for Technical University of Denmark

Thu, 11/02/2017 - 08:17

SHENZHEN, China, Nov. 2, 2017 — For almost two centuries DTU, Technical University of Denmark, has been dedicated to fulfilling the vision of H.C. Orsted – the father of electromagnetism – who founded the university in 1829 to develop and create value using the natural sciences and the technical sciences to benefit society. Today, DTU is ranked as one of the foremost technical universities in Europe.

High-Performance Computing Propels Materials Research

DTU promotes promising fields of research within the technical and the natural sciences, especially based on usefulness to society, relevance to business and sustainability. DTU focuses on basic science that has significant challenges and clear application prospects, from atomic-scale materials analysis to quantum physics and renewable energy. As the material application environment becomes increasingly complex, laboratory research for materials performance analysis has become even more challenging.

DTU aims to understand the nature of materials by developing electron structural theory, and design new functional nanostructures through new-found insights. These studies require the analysis of the structure, strength, and characteristics of new materials, involving intensive, complex numerical computation and simulation tests on material and energy. This will produce a vast number of computational data. Therefore, High-Performance Computing (HPC) resources that can accelerate performance modeling and solving are particularly important to research in this field.

In order to speed up the process from discovery to application of new materials and maintain a leading edge in research, DTU plans to expand and upgrade its supercomputing cluster, Niflheim, which is deployed at the Computational Atomic-scale Materials Design (CAMD) Center.

Combining the Best of Both Worlds: Huawei X6800 High-Density Server and Intel OPA Network

The existing Niflheim cluster at DTU was built from 2009 to 2015, and was capable of a peak computing capability of only 73 TFLOPS. The cluster was equipped with previous generation and even earlier computing product hardware. The oldest products had limited processor performance, small memory capacity, with low-bandwidth but high-latency computing network. The old cluster was failing to meet the growing demands of computing-intensive simulation tests. As a result, the cluster became a bottleneck since the CAMD center needed research efficiency improvements.

DTU wanted to deploy a new supercomputing system to give the Niflheim cluster a boost in computing resources and performance, and meanwhile also prepare the cluster for future technology evolution as well as cluster-scale expansion. DTU has carefully studied various solutions in terms of overall performance, product quality, and service capabilities, and through an EU tender finally selected Huawei and Intel as the vendors to help the university build a new-generation computing cluster with their innovative technologies and computing products.

Solution Highlights

Supreme Performance, Leading Computing Efficiency:

Nodes configured with Intel® Xeon® E5-2600 v4 series processors, up to 845 GFLOPS compute power per node;

Nodes configured with 256 GB DIMMs and 240 GB SSDs, eliminates I/O bottlenecks, and improves data processing efficiency with high-speed data caching;

Leverages the Intel® Omni-Path Architecture (OPA) to build a two-layer fat-tree fabric, delivers bandwidth of up to 100 Gbit/s, and end-to-end latency as low as 910 ns;

Power Supply Units (PSUs) and fan modules shared by multiple nodes, enhanced with Huawei’s Dynamic Energy Management Technology (DEMT) to lower system energy consumption by over 10%.

High-Density Deployment, Easy to Manage and Expand:

4U chassis configured with eight 2-socket compute nodes, delivers computing density twice that of traditional 1U rack servers, significantly improves rack space utilization;

Supports aggregated management network port for unified management, meanwhile reduces cable connections;

Adopts a modular design, and supports hot swap for all key components, greatly improves Operations and Maintenance (O&M) efficiency.

New-Generation Niflheim Cluster Expedites New Material Discovery and Application

The new-generation Niflheim cluster went live in Dec. 2016. The new cluster helps more researchers carry out research and analysis on new materials and new energy, but also provides a great leap in feedback speeds of test results. It has enabled new levels of scientific research progress and strength, helping DTU generate new innovation capabilities in the field of material analysis.

The Niflheim cluster delivers a computing power of up to 225 TFLOPS, which is three times the level of the original system;

Substantially shorten the materials’ analysis time, enabling researchers to discover and apply new materials more quickly;

With flexible expandability, the cluster can seamlessly expand up to 112 nodes without requiring additional new cabinets.

Source: Huawei

The post Huawei Partners with Intel to Build a Supercomputing Cluster for Technical University of Denmark appeared first on HPCwire.