Tagged: insideHPC Toggle Comment Threads | Keyboard Shortcuts

  • richardmitnick 11:12 am on April 23, 2019 Permalink | Reply
    Tags: DiRAC is the integrated supercomputing facility for theoretical modeling and HPC-based research in particle physics and astrophysics cosmology and nuclear physics all areas in which the UK is world-le, insideHPC,   

    From insideHPC: “40 Powers of 10 – Simulating the Universe with the DiRAC HPC Facility” 

    From insideHPC

    DiRAC is the UK’s integrated supercomputing facility for theoretical modelling and HPC-based research in particle physics, astronomy and cosmology.


    49 minutes, worth your time
    In this video from the Swiss HPC Conference, Mark Wilkinson presents: 40 Powers of 10 – Simulating the Universe with the DiRAC HPC Facility.

    2
    Dr. Mark Wilkinson is the Project Director at DiRAC.

    “DiRAC is the integrated supercomputing facility for theoretical modeling and HPC-based research in particle physics, and astrophysics, cosmology, and nuclear physics, all areas in which the UK is world-leading. DiRAC provides a variety of compute resources, matching machine architecture to the algorithm design and requirements of the research problems to be solved. As a single federated Facility, DiRAC allows more effective and efficient use of computing resources, supporting the delivery of the science programs across the STFC research communities. It provides a common training and consultation framework and, crucially, provides critical mass and a coordinating structure for both small- and large-scale cross-discipline science projects, the technical support needed to run and develop a distributed HPC service, and a pool of expertise to support knowledge transfer and industrial partnership projects. The on-going development and sharing of best-practice for the delivery of productive, national HPC services with DiRAC enables STFC researchers to produce world-leading science across the entire STFC science theory program.”

    3

    Dr. Mark Wilkinson is the Project Director at DiRAC. He obtained his BA and MSc in Theoretical Physics at Trinity College Dublin and a DPhil in Theoretical Astronomy at the University of Oxford. Between 2000 and 2006, he was a post-doc at the Institute of Astronomy, Cambridge. He subsequently moved to the University of Leicester to take up a Royal Society University Research Fellow. I am currently a Reader in the Theoretical Astrophysics Group of the Dept. of Physics & Astronomy.

    See the full article here .

    five-ways-keep-your-child-safe-school-shootings

    Please help promote STEM in your local schools.

    Stem Education Coalition

    Founded on December 28, 2006, insideHPC is a blog that distills news and events in the world of HPC and presents them in bite-sized nuggets of helpfulness as a resource for supercomputing professionals. As one reader said, we’re sifting through all the news so you don’t have to!

    If you would like to contact me with suggestions, comments, corrections, errors or new company announcements, please send me an email at rich@insidehpc.com. Or you can send me mail at:

    insideHPC
    2825 NW Upshur
    Suite G
    Portland, OR 97239

    Phone: (503) 877-5048

     
  • richardmitnick 8:21 am on April 15, 2019 Permalink | Reply
    Tags: "Wave Computing Launches TritonAI 64 Platform for High-Speed Inferencing", insideHPC   

    From insideHPC: “Wave Computing Launches TritonAI 64 Platform for High-Speed Inferencing” 

    From insideHPC

    April 14, 2019

    1

    Today AI startup Wave Computing announced its new TritonAI 64 platform, which integrates a triad of powerful technologies into a single, future-proof intellectual property (IP) licensable solution. Wave’s TritonAI 64 platform delivers 8-to-32-bit integer-based support for high-performance AI inferencing at the edge now, with bfloat16 and 32-bit floating point-based support for edge training in the future.

    “Wave Computing is achieving another industry first by delivering a licensable IP platform that enables both AI inferencing and training at the edge,” said Derek Meyer, Chief Executive Officer of Wave Computing. “The tremendous growth of edge-based AI use cases is exacerbating the challenges of SoC designers who continue to struggle with legacy IP products that were not designed for efficient AI processing. Our TritonAI solution provides them with the investment protection of a programmable platform that can scale to support the AI applications of both today and tomorrow. TritonAI 64 enhances our overall AI offerings that span datacenter to edge and is another company milestone enabled by our acquisition of MIPS last year.”

    Wave’s TritonAI 64 platform is an industry-first solution, enabling customers to address a broad range of AI use cases with a single platform. The platform delivers efficient edge inferencing and training performance to support today’s AI algorithms, while providing customers with the flexibility to future-proof their investment for emerging AI algorithms. Features of the TritonAI 64 platform include a leading-edge MIPS® 64-bit SIMD engine that is integrated with Wave’s unique approach to dataflow and tensor-based configurable technology. Additional features inclu­­de access to Wave’s MIPS integrated developer environment (IDE), as well as a Linux-based TensorFlow programming environment.

    The global market for AI products is projected to dramatically increase to over $170B by 2025, according to technology analyst firm Tractica. The total addressable market (TAM) for AI at the edge comprises over $100B of this market and is being driven primarily by the needs for more efficient inferencing, and new AI workloads and use cases, as well as the need for training at the edge.


    In this video from the Hot Chips conference, Chris Nicol from Wave Computing presents: A Dataflow Processing Chip for Training Deep Neural Networks.
    1:10:22

    Details of Wave’s TritonAI 64 Platform:

    MIPS 64-bit + SIMD Technology: Offering an open instruction set architecture (MIPS Open), coupled with a mature integrated development environment (IDE), provides an ideal software platform for developing AI applications, stacks and use cases. The MIPS IP subsystem in the TritonAI 64 platform enables SoCs to be configured with up to six MIPS 64 CPUs, each with up to four hardware-threads. The MIPS subsystem hosts the execution of Google’s TensorFlow framework on a debian-based Linux operating system, enabling the development of both inferencing and edge learning applications. Additional AI frameworks such as Caffe2, can be ported to the MIPS subsystem, as well as support a wide variety of AI networks using ONNX conversion.
    WaveTensor Technology: The WaveTensor subsystem can scale up to a PetaOP of 8-bit integer operations on a single core instantiation by combining extensible slices of 4×4 or 8×8 kernel matrix multiplier engines for the highly efficient execution of today’s key Convolutional Neural Network (CNN) algorithms. The CNN execution performance can scale up to 8 TOPS/watt and over 10 TOPS/mm2 in industry standard 7nm process nodes with libraries using typical voltage and processes.
    WaveFlow Technology: Wave Computing’s highly flexible, linearly scalable fabric is adaptable for any number of complex AI algorithms, as well as conventional signal processing and vision algorithms. The WaveFlow subsystem features low latency, single batch size AI network execution and reconfigurability to address concurrent AI network execution. This patented WaveFlow architecture also supports algorithm execution without intervention or support from the MIPS subsystem.

    See the full article here .

    five-ways-keep-your-child-safe-school-shootings

    Please help promote STEM in your local schools.

    Stem Education Coalition

    Founded on December 28, 2006, insideHPC is a blog that distills news and events in the world of HPC and presents them in bite-sized nuggets of helpfulness as a resource for supercomputing professionals. As one reader said, we’re sifting through all the news so you don’t have to!

    If you would like to contact me with suggestions, comments, corrections, errors or new company announcements, please send me an email at rich@insidehpc.com. Or you can send me mail at:

    insideHPC
    2825 NW Upshur
    Suite G
    Portland, OR 97239

    Phone: (503) 877-5048

     
  • richardmitnick 11:15 am on April 13, 2019 Permalink | Reply
    Tags: , , insideHPC, LLNL Penguin Computing Corona AMD EPYC Radeon Instinct Cluster, ,   

    From insideHPC: “AMD Powers Corona Cluster for HPC Analytics at Livermore” 

    From insideHPC

    April 12, 2019
    Rich Brueckner

    Lawrence Livermore National Lab has deployed a 170-node HPC cluster from Penguin Computing. Based on AMD EPYC processors and Radeon Instinct GPUs, the new Corona cluster will be used to support the NNSA Advanced Simulation and Computing (ASC) program in an unclassified site dedicated to partnerships with American industry.

    1
    The Corona cluster is comprised of AMD EPYC processors, AMD Radeon Instinct GPUs connected with Mellanox HDR 200 Gigabit InfiniBand.

    In searching for a commercial processor that could handle the demands of HPC and data analytics, Matt Leininger, Deputy for Advanced Technology Projects, LLNL, said several factors influence the choice of CPU including single core performance, the number of cores, and the memory performance per socket. All these factors drove LLNL to seriously consider the EPYC processor.

    “Our simulations require a balance of memory and compute capabilities. The number of high-performance memory channels and CPU cores on each AMD EPYC socket are a solution for our mission needs.” he said.

    The lab’s latest HPC cluster deployment—named Corona—is built with AMD EPYC CPUs and Radeon Instinct MI25 GPUs. “We are excited to have these high-end products and to apply them to our challenging HPC simulations,” said Leininger.

    Simulations requiring 100’s of petaFLOPS (quadrillions of floating-point operations per second) speed are run on the largest supercomputers at LLNL, which are among the fastest in the world. Supporting the larger systems are the Commodity Technology Systems (CTS), what Leininger calls “everyday workhorses” serving the LLNL user community.

    The new Corona cluster will bring new levels of machine learning capabilities to the CTS resources. The integration of HPC simulation and machine learning into a cognitive simulation capability is an active area of research at LLNL.

    “Coupling large scale deep learning with our traditional scientific simulation workload will allow us to dramatically increase scientific output and utilize our HPC resources more efficiently,” said LLNL Informatics Group leader and computer scientist Brian Van Essen. “These new Deep Learning enabled HPC systems are critical as we develop new machine learning algorithms and architectures that are optimized for scientific computing.”

    The computing platform is Penguin Computing’s XO1114GT platform, with nodes connected by Mellanox HDR InfiniBand networking technology.

    “We have folks thinking about what they can pull off on this machine that they couldn’t have done before,” Leininger said.

    CPU/GPU Powering Machine Learning in HPC

    “We’ve been working to understand how to enable HPC simulation using GPUs and also using machine learning in combination with HPC to solve some of our most challenging scientific problems,” Leininger said. “Even as we do more of our computing on GPUs, many of our codes have serial aspects that need really good single core performance. That lines up well with AMD EPYC.”

    The EPYC processor-based Corona cluster will help LLNL use machine learning to conduct its simulations more efficiently with an active-learning approach, called Cognitive Simulation, that can be used to optimize solutions with a significant reduction in compute requirements. Leininger explained that multi-physics simulations, which include a significant number of modeling and calculations around hydrodynamic and materials problems important to NNSA, are the lab’s most complicated. These analytic simulations produce a range of parameter space results that are used to construct error bars which depict uncertainty levels that must be understood and reduced.

    “We are looking to use some machine learning techniques where the machine would figure out how much of the parameter space we really need to explore or what part of it we need to explore more than others,” Van Essen said.

    Using EPYC-powered servers with the Radeon Instinct MI25 for machine learning, LLNL will be able to determine exactly where to explore further in order to detect what component is driving the majority of the error bars and significantly reduce time on task to achieve better science.

    See the full article here .

    five-ways-keep-your-child-safe-school-shootings

    Please help promote STEM in your local schools.

    Stem Education Coalition

    Founded on December 28, 2006, insideHPC is a blog that distills news and events in the world of HPC and presents them in bite-sized nuggets of helpfulness as a resource for supercomputing professionals. As one reader said, we’re sifting through all the news so you don’t have to!

    If you would like to contact me with suggestions, comments, corrections, errors or new company announcements, please send me an email at rich@insidehpc.com. Or you can send me mail at:

    insideHPC
    2825 NW Upshur
    Suite G
    Portland, OR 97239

    Phone: (503) 877-5048

     
  • richardmitnick 9:24 am on April 13, 2019 Permalink | Reply
    Tags: A new photonic switch built with more than 50000 microscopic “light switches”, , Each switch directs one of 240 tiny beams of light to either make a right turn when the switch is on or to pass straight through when the switch is off, insideHPC, , Photolithography, Server networks could be connected by optical fibers with photonic switches acting as the traffic cops Wu said, This could one day revolutionize how information travels through data centers and high-performance supercomputers that are used for artificial intelligence and other data-intensive applications.,   

    From insideHPC: “Berkeley Engineers build World’s Fastest Optical Switch Arrays” 

    From insideHPC

    April 12, 2019

    Engineers at the University of California, Berkeley have built a new photonic switch that can control the direction of light passing through optical fibers faster and more efficiently than ever.

    1
    The photonic switch is built with more than 50,000 microscopic “light switches” etched into a silicon wafer. (Younghee Lee graphic)

    This optical “traffic cop” could one day revolutionize how information travels through data centers and high-performance supercomputers that are used for artificial intelligence and other data-intensive applications.

    The photonic switch is built with more than 50,000 microscopic “light switches,” each of which directs one of 240 tiny beams of light to either make a right turn when the switch is on, or to pass straight through when the switch is off. The 240-by-240 array of switches is etched into a silicon wafer and covers an area only slightly larger than a postage stamp.

    “For the first time in a silicon switch, we are approaching the large switches that people can only build using bulk optics,” said Ming Wu, professor of electrical engineering and computer sciences at UC Berkeley and senior author of the paper, which appeared online April 11 in the journal Optica. “Our switches are not only large, but they are 10,000 times faster, so we can switch data networks in interesting ways that not many people have thought about.”

    Currently, the only photonic switches that can control hundreds of light beams at once are built with mirrors or lenses that must be physically turned to switch the direction of light. Each turn takes about one-tenth of a second to complete, which is eons compared to electronic data transfer rates. The new photonic switch is built using tiny integrated silicon structures that can switch on and off in a fraction of a microsecond, approaching the speed necessary for use in high-speed data networks.

    Traffic cops on the information highway

    Data centers — where our photos, videos and documents saved in the cloud are stored — are composed of hundreds of thousands of servers that are constantly sending information back and forth. Electrical switches act as traffic cops, making sure that information sent from one server reaches the target server and doesn’t get lost along the way.

    But as data transfer rates continue to grow, we are reaching the limits of what electrical switches can handle, Wu said.

    “Electrical switches generate so much heat, so even though we could cram more transistors onto a switch, the heat they generate is starting to pose certain limits,” he said. “Industry expects to continue the trend for maybe two more generations and, after that, something more fundamental has to change. Some people are thinking optics can help.”

    Server networks could instead be connected by optical fibers, with photonic switches acting as the traffic cops, Wu said. Photonic switches require very little power and don’t generate any heat, so they don’t face the same limitations as electrical switches. However, current photonic switches cannot accommodate as many connections and also are plagued by signal loss — essentially “dimming” the light as it passes through the switch — which makes it hard to read the encoded data once it reaches its destination.

    In the new photonic switch, beams of light travel through a crisscrossing array of nanometer-thin channels until they reach these individual light switches, each of which is built like a microscopic freeway overpass. When the switch is off, the light travels straight through the channel. Applying a voltage turns the switch on, lowering a ramp that directs the light into a higher channel, which turns it 90 degrees. Another ramp lowers the light back into a perpendicular channel.

    “It’s literally like a freeway ramp,” Wu said. “All of the light goes up, makes a 90-degree turn and then goes back down. And this is a very efficient process, more efficient than what everybody else is doing on silicon photonics. It is this mechanism that allows us to make lower-loss switches.”

    The team uses a technique called photolithography to etch the switching structures into silicon wafers. The researchers can currently make structures in a 240-by-240 array — 240 light inputs and 240 light outputs — with limited light loss, making it the largest silicon-based switch ever reported. They are working on perfecting their manufacturing technique to create even bigger switches.

    “Larger switches that use bulk optics are commercially available, but they are very slow, so they are usable in a network that you don’t change too frequently,” Wu said. “Now, computers work very fast, so if you want to keep up with the computer speed, you need much faster switch response. Our switch is the same size, but much faster, so it will enable new functions in data center networks.”

    Co-lead authors on the paper are Tae Joon Seok of the Gwangju Institute of Science and Technology and Kyungmok Kwon, a postdoctoral researcher and Bakar Innovation Fellow at UC Berkeley. Other co-authors are Johannes Henriksson and Jianheng Luo of UC Berkeley.

    See the full article here .

    five-ways-keep-your-child-safe-school-shootings

    Please help promote STEM in your local schools.

    Stem Education Coalition

    Founded on December 28, 2006, insideHPC is a blog that distills news and events in the world of HPC and presents them in bite-sized nuggets of helpfulness as a resource for supercomputing professionals. As one reader said, we’re sifting through all the news so you don’t have to!

    If you would like to contact me with suggestions, comments, corrections, errors or new company announcements, please send me an email at rich@insidehpc.com. Or you can send me mail at:

    insideHPC
    2825 NW Upshur
    Suite G
    Portland, OR 97239

    Phone: (503) 877-5048

     
  • richardmitnick 12:23 pm on April 12, 2019 Permalink | Reply
    Tags: "Scaling Deep Learning for Scientific Workloads on the #1 Summit Supercomputer", insideHPC, , , ORNL Cray XK7 Titan Supercomputer once the fastest in the world now No.9 on the TOP500, ORNL IBM AC922 SUMMIT supercomputer No.1 on the TOP500   

    From insideHPC: “Scaling Deep Learning for Scientific Workloads on the #1 Summit Supercomputer” 

    From insideHPC

    April 11, 2019
    Rich Brueckner


    In this video from GTC 2018, Jack Wells from ORNL presents: Scaling Deep Learning for Scientific Workloads on Summit.

    2
    Jack Wells is the Director of Science for the Oak Ridge Leadership Computing Facility (OLCF).

    “HPC centers have been traditionally configured for simulation workloads, but deep learning has been increasingly applied alongside simulation on scientific datasets. These frameworks do not always fit well with job schedulers, large parallel file systems, and MPI backends. We’ll discuss examples of how deep learning workflows are being deployed on next-generation systems at the Oak Ridge Leadership Computing Facility. We’ll share benchmarks between native compiled versus containers on Power systems, like Summit, as well as best practices for deploying learning and models on HPC resources on scientific workflows.”

    The biggest problems in science require supercomputers of unprecedented capability. That’s why the US Department of Energy’s Oak Ridge National Laboratory (ORNL) launched Summit, a system 8 times more powerful than ORNL’s previous top-ranked system Titan. Summit is providing scientists with incredible computing power to solve challenges in energy, artificial intelligence, human health, and other research areas, that were simply out of reach until now. These discoveries will help shape our understanding of the universe, bolster US economic competitiveness, and contribute to a better future.

    ORNL IBM AC922 SUMMIT supercomputer, No.1 on the TOP500. Credit: Carlos Jones, Oak Ridge National Laboratory/U.S. Dept. of Energy

    Summit Specifications:
    Application Performance: 200 PF (currently #1 on the TOP500)
    Number of Nodes: 4,608
    Node performance: 42 TF
    Memory per Node: 512 GB DDR4 + 96 GB HBM2
    NV memory per Node: 1600 GB
    Total System Memory: >10 PB DDR4 + HBM2 + Non-volatile
    Processors:
    2 IBM POWER9 9,216 CPUs
    6 NVIDIA Volta 27,648 GPUs

    File System: 250 PB, 2.5 TB/s, GPFS
    Power Consumption: 13 MW
    Interconnect: Mellanox EDR 100G InfiniBand
    Operating System: Red Hat Enterprise Linux (RHEL) version 7.4

    Jack Wells is the Director of Science for the Oak Ridge Leadership Computing Facility (OLCF), a DOE Office of Science national user facility, and the Titan supercomputer, located at Oak Ridge National Laboratory (ORNL).

    ORNL Cray XK7 Titan Supercomputer, once the fastest in the world, now No.9 on the TOP500.

    Wells is responsible for the scientific outcomes of the OLCF’s user programs. Wells has previously lead both ORNL’s Computational Materials Sciences group in the Computer Science and Mathematics Division and the Nanomaterials Theory Institute in the Center for Nanophase Materials Sciences. Prior to joining ORNL as a Wigner Fellow in 1997, Wells was a postdoctoral fellow within the Institute for Theoretical Atomic and Molecular Physics at the Harvard-Smithsonian Center for Astrophysics. Wells has a Ph.D. in physics from Vanderbilt University, and has authored or co-authored over 100 scientific papers and edited 1 book, spanning nanoscience, materials science and engineering, nuclear and atomic physics computational science, applied mathematics, and novel analytics measuring the impact of scientific publications.

    See the full article here .

    five-ways-keep-your-child-safe-school-shootings

    Please help promote STEM in your local schools.

    Stem Education Coalition

    Founded on December 28, 2006, insideHPC is a blog that distills news and events in the world of HPC and presents them in bite-sized nuggets of helpfulness as a resource for supercomputing professionals. As one reader said, we’re sifting through all the news so you don’t have to!

    If you would like to contact me with suggestions, comments, corrections, errors or new company announcements, please send me an email at rich@insidehpc.com. Or you can send me mail at:

    insideHPC
    2825 NW Upshur
    Suite G
    Portland, OR 97239

    Phone: (503) 877-5048

     
  • richardmitnick 10:02 am on April 8, 2019 Permalink | Reply
    Tags: "Hyperion Research: HPC Server Market Beat Forecast in 2018", , , , insideHPC, QView,   

    From insideHPC: “Hyperion Research: HPC Server Market Beat Forecast in 2018” 

    From insideHPC

    April 7, 2019

    1
    Steve Conway is Senior Research Vice President at Hyperion Research.

    Hyperion Research has released their latest High-Performance Technical Server QView, a comprehensive report on the state of the HPC Market. The QView presents the HPC market from various perspectives, including competitive segment, vendor, cluster versus non-cluster, geography, and operating system. It also contains detailed revenue and shipment information by HPC models.

    “Worldwide factory revenue for the high-performance computing (HPC) technical server market jumped for the full-year 2018 by 15.6% over the exceptionally strong 2017 calendar year period, from $11.9 billion to $13.7 billion. The 2018 result was 6.2% higher than Hyperion Research’s forecast $12.9 billion figure for the year.”

    For the fourth quarter of 2018, the market grew by 10.1% to reach $3.7 billion, up from $3.5 billion in the same period of 2017, according to the newly released Hyperion Research Worldwide High-Performance Technical Server QView.
    Hyperion Research forecasts that revenue for HPC server systems will grow at a 7.8% CAGR to reach $20.0 billion in 2023.
    Leading the charge in year-over-year fourth-quarter growth was the Supercomputers segment for HPC systems sold for $500,000 and up. Revenue in this segment climbed 31.0% to $1.6 billion, compared with $1.2 billion in the prior-year fourth quarter, and accounted for 41.9% of fourth-quarter 2018 HPC server revenue.
    The Divisional segment for systems sold for $250,000 to $499,000 grew by 12.9%, from $587 million in the 2017 fourth quarter to $662 million in the 2018 fourth quarter, accounting for 17.8% of fourth-quarter 2018 revenue.
    The Departmental segment for systems sold for $100,000 to $249,000 declined by 6.4% in the 2018 fourth quarter to $1.0 billion from the strong 2017 fourth-quarter total of $1.1 billion. The 2018 figure represented 27.0% of fourth-quarter HPC server revenue.
    Similarly, the Workgroup segment for HPC systems sold for under $100,000 dipped 18.7%, from $607 million in the 2017 quarter to $493 million in fourth-quarter 2018, representing 13.3% of fourth-quarter 2018 HPC server revenue.

    “Fourth-quarter and calendar-year 2018 HPC server revenue grew even faster than we forecast, helped by a robust U.S. and global economy,” said Earl Joseph, CEO of Hyperion Research. “We predict continued healthy growth through 2023, with some banner years for the Supercomputers segment as pre-exascale and exascale supercomputers are deployed across the world.”

    According to Steve Conway, Hyperion Research senior vice president of research, “fourth-quarter revenue also benefited from HPC’s crucial role at the forefront of AI research and from the growing adoption of HPC servers by enterprise data centers to accelerate business operations.”

    Vendor Highlights for Fourth-Quarter 2018

    Hewlett Packard Enterprise and Dell EMC remained the global market leaders by capturing, respectively, 34.9% and 22.7% of worldwide HPC server system revenue.
    Lenovo was the third-place finisher with a 6.8% share.
    The largest year-over-year fourth-quarter market share gainers were Penguin Computing (58.0%), Fujitsu (32.9%), Sugon (28.0%) and Dell EMC (27.0%).


    In this video from the HPC User Forumin Santa Fe, Earl Joseph from Hyperion Research presents an HPC Market Update.

    See the full article here .

    five-ways-keep-your-child-safe-school-shootings

    Please help promote STEM in your local schools.

    Stem Education Coalition

    Founded on December 28, 2006, insideHPC is a blog that distills news and events in the world of HPC and presents them in bite-sized nuggets of helpfulness as a resource for supercomputing professionals. As one reader said, we’re sifting through all the news so you don’t have to!

    If you would like to contact me with suggestions, comments, corrections, errors or new company announcements, please send me an email at rich@insidehpc.com. Or you can send me mail at:

    insideHPC
    2825 NW Upshur
    Suite G
    Portland, OR 97239

    Phone: (503) 877-5048

     
  • richardmitnick 6:00 pm on April 3, 2019 Permalink | Reply
    Tags: insideHPC, Pittsburgh Supercompute Center, Thorny Flat High Performance Computer Cluster   

    From insideHPC: “Thorny Flat Supercomputer comes to West Virginia University” 

    From insideHPC

    Today West Virginia University announced of one of the state’s most powerful computer clusters to help power research and innovation statewide.

    “The Thorny Flat High Performance Computer Cluster [no specs available,no rank], named after the state’s second highest peak, joins the Spruce Knob cluster [no image or specs available] as resources. With 1,000 times more computing power than a desktop computer, the Thorny Flat cluster could benefit a variety research: forest hydrology; genetic studies; forensic chemistry of firearms; modeling of solar-to-chemical energy harvesting; and design and discovery of new materials.”

    Pittsburgh Super Computer Center Thorny Flat Supercomputer – no better image available nor any specs

    Thorny Flat is the latest step in an ongoing regional partnership with the Pittsburgh Supercomputing Center, which has contributed greatly to the Morgantown-Pittsburgh corridor’s status as a hub for technological and scientific progress.

    The system will be housed in PSC’s machine room, and the center will provide routine maintenance and support for scientists using it.

    The cluster was created by WVU’s Research Computing team in Information Technology Services with a nearly $1 million grant from the National Science Foundation Office of Advanced Cyberinfrastructure. Thorny Flat’s creation also underscores WVU’s commitment to ground-breaking research, which is being highlighted this week in the institution’s inaugural Research Week.

    In his latest State of the University address, WVU President Gordon Gee challenged the community to innovate so West Virginia can remain vibrant and competitive. “In a state that no longer manufactures products the way it used to, West Virginia University’s faculty, staff and students have no choice but to manufacture transformation,” he said. “We must pioneer progress.”

    Blake Mertz, assistant professor of chemistry in WVU’s Eberly College of Arts and Sciences, was the principal investigator behind the NSF grant. Mertz’s lab uses molecular dynamics simulations to model systems of interest in both biomedical and alternative energy research.

    In particular, Mertz is studying a peptide (a small protein) that could be used for targeted drug deliveries in cancer patients. Thorny Flat acts as a sort of computational microscope for observing key interactions between the peptide and the cell membrane surface. Mertz said these simulations can speed discoveries about the peptide, saving not only time and money in the lab, but also lives in the clinic.

    Also from Eberly, WVU physicist Zachariah Etienne will use Thorny Flat for simulations that model the physical processes involved when neutron stars collide.

    “Neutron stars are not simply large balls of neutrons, and we cannot create neutron star matter in laboratories,” Etienne said. The only way to understand how such matter behaves is to compare observations with computer simulations.”

    This kind of research “has the potential to revolutionize our current scientific understanding,” Etienne said. “Historically, such revolutions have led to incredible new technologies that help us diagnose and cure disease, and generally make our modern lives far safer and more comfortable than our predecessors.”

    Thorny Flat, which will become available Tuesday (April 9), can accommodate a wide range of users, offering free access to researchers with on-demand computational projects and first-time HPC users. Researchers with more intensive, long-term computational needs also can purchase separate, dedicated time and resources.

    With 108 total nodes and 4,208 cores, Thorny Flat has a peak performance of 428 TeraFLOPS. It also features 21 GPUs for AI research.

    See the full article here .

    five-ways-keep-your-child-safe-school-shootings

    Please help promote STEM in your local schools.

    Stem Education Coalition

    Founded on December 28, 2006, insideHPC is a blog that distills news and events in the world of HPC and presents them in bite-sized nuggets of helpfulness as a resource for supercomputing professionals. As one reader said, we’re sifting through all the news so you don’t have to!

    If you would like to contact me with suggestions, comments, corrections, errors or new company announcements, please send me an email at rich@insidehpc.com. Or you can send me mail at:

    insideHPC
    2825 NW Upshur
    Suite G
    Portland, OR 97239

    Phone: (503) 877-5048

     
  • richardmitnick 5:57 pm on April 2, 2019 Permalink | Reply
    Tags: insideHPC, Piz Daint supercomputer to be used by CERN as Tier 2 system, Replaces the Phoenix supercomputer cluster which will be decommissioned   

    From insideHPC: “Piz Daint Supercomputer to Power LHC Computing Grid” 

    From insideHPC

    April 2, 2019

    The fastest supercomputer in Europe will soon join the WLHC Grid. Housed at CSCS in Switzerland, the Piz Daint supercomputer be used for data analysis from Large Hadron Collider (LHC) experiments.

    Cray Piz Daint Cray XC50/XC40 supercomputer of the Swiss National Supercomputing Center (CSCS)

    Until now, the ATLAS, CMS and LHCb particle detectors delivered their data to “Phoenix” system for analysis and comparison with the results of previous simulations.

    A unique approach

    For the first time, the Worldwide LHC Computing Grid (WLCG) will integrate one of the world’s most powerful supercomputers – which is available for general research – to handle the functions of what is known as a Tier 2 system. Until now, this role has been handled exclusively by dedicated clusters distributed around the world in the WLCG. Based on their performance and characteristics, the clusters are categorized on a scale ranging from Tier 0 systems – such as those available only at the CERN data centre in Geneva and the Wigner Research Centre for Physics in Budapest – to smaller Tier 3 systems.

    “CSCS has been operating the “Phoenix” cluster, which accounts for about 3% of the WLCG’s capacity, as a Tier 2 system for about 12 years.

    Phoenix Intel AMD supercomputer cluster

    The fact that “Piz Daint” will now be able to handle the data analysis of one of the world’s most data-intensive projects in addition to its daily tasks is thanks to a long-term partnership between CSCS and CHIPP. The project participants have been working on its implementation since 2014.

    A host of benefits – and challenges

    In addition to the advantages classic supercomputers offer physicists, such as scalability, high performance and efficiency, other properties have taken some getting used to for the high-energy physicists. “The integration of ‘Piz Daint’ with the high-energy physics data processing environments has been a highly laborious process that started with the integration of a previous generation supercomputer at CSCS,” says Gianfranco Sciacca, who is part of the ATLAS experiment. “The data access patterns and rates of our workflows are not typical of a supercomputer environment. In addition, for some workflows, certain resource requirements exceed what a general-purpose supercomputer typically provides; a lot of tuning therefore needs to be put in place.”

    “It was also a challenge for CSCS: “Such a move away from the status quo, with all five parties – CHIPP, ATLAS, CMS, LHCb and CSCS – sharing the same view, both on conceptual and practical grounds, was probably the biggest challenge that we have had to face in this project,” says Pablo Fernandez, Service and Business Manager at CSCS. Fernandez coordinated a whole range of activities during the project.”

    Migrating tasks from “Phoenix” to “Piz Daint” took place gradually over two years. Only at the end of 2017, when enough comparative data was available to make an informed decision, did CSCS and CHIPP decide to use “Piz Daint” for future data analysis. “Piz Daint” has been used actively for simulation and data analysis in LHC experiments since April 2018. In April 2019, the transfer will finally be complete and “Phoenix” will be decommissioned. Meanwhile, the project participants have succeeded in testing the externalization of the Tier 0 system onto “Piz Daint” and have provided CERN with direct computing capacity for highly complex data processing. The project participants look back at what they have achieved with a certain degree of awe: “The project was far more complex to manage than any technical limitation or problem we have had so far,” says Miguel Gila, HPC System Engineer at CSCS.

    More work on the project to come

    The comparison between “Piz Daint” and “Phoenix” has shown that both systems perform their functions with similar efficiency, but “Piz Daint” is slightly more cost-effective. According to the project participants, however, the CHIPP community will use only the computer nodes consisting of two CPUs for the time being, and not all the full benefits of “Piz Daint”. In future, it may be possible to use hybrid computer nodes consisting of one CPU and one GPU to run calculations much more efficiently. A big advantage for the researchers is that they can tap into much more computing power at short notice – at the “push of a button” – if needed.

    The current upgrade of the LHC to the High-Luminosity LHC, which will be online from 2025 to 2034 and enable breakthroughs in particle physics, requires up to 50 times more computing power for data analysis and simulations. However, the high energy physicists predict that with the current computing models, a significant shortfall of resources will occur. This will require various software and hardware optimizations. Since HPC systems such as “Piz Daint” already have optimised hardware, the researchers are confident that it will be possible to make savings. They expect the project to serve as a model for WLCG, as it has opened doors, for instance, in terms of new computing models and software optimization.

    See the full article here .

    five-ways-keep-your-child-safe-school-shootings

    Please help promote STEM in your local schools.

    Stem Education Coalition

    Founded on December 28, 2006, insideHPC is a blog that distills news and events in the world of HPC and presents them in bite-sized nuggets of helpfulness as a resource for supercomputing professionals. As one reader said, we’re sifting through all the news so you don’t have to!

    If you would like to contact me with suggestions, comments, corrections, errors or new company announcements, please send me an email at rich@insidehpc.com. Or you can send me mail at:

    insideHPC
    2825 NW Upshur
    Suite G
    Portland, OR 97239

    Phone: (503) 877-5048

     
  • richardmitnick 8:52 pm on March 28, 2019 Permalink | Reply
    Tags: , , , , , insideHPC, , , ,   

    From insideHPC: “Nor-Tech Powers LIGO and IceCube Nobel-Physics Prize-Winning Projects” 

    From insideHPC

    March 28, 2019

    Today HPC integrator Nor-Tech announced participation in two recent Nobel Physics Prize-Winning projects. The company’s HPC gear will help power the Laser Interferometer Gravitational-Wave Observatory (LIGO) project as well as the IceCube neutrino detection experiment.


    VIRGO Gravitational Wave interferometer, near Pisa, Italy

    Caltech/MIT Advanced aLigo Hanford, WA, USA installation


    Caltech/MIT Advanced aLigo detector installation Livingston, LA, USA

    Cornell SXS, the Simulating eXtreme Spacetimes (SXS) project

    Gravitational waves. Credit: MPI for Gravitational Physics/W.Benger

    Gravity is talking. Lisa will listen. Dialogos of Eide

    ESA/eLISA the future of gravitational wave research

    Localizations of gravitational-wave signals detected by LIGO in 2015 (GW150914, LVT151012, GW151226, GW170104), more recently, by the LIGO-Virgo network (GW170814, GW170817). After Virgo came online in August 2018


    Skymap showing how adding Virgo to LIGO helps in reducing the size of the source-likely region in the sky. (Credit: Giuseppe Greco (Virgo Urbino group)

    U Wisconsin IceCube neutrino observatory

    U Wisconsin ICECUBE neutrino detector at the South Pole

    U Wisconsin IceCube experiment at the South Pole



    U Wisconsin ICECUBE neutrino detector at the South Pole


    IceCube Gen-2 DeepCore PINGU


    IceCube reveals interesting high-energy neutrino events

    “We are excited about the amazing discoveries these enhanced detectors will reveal,” said Nor-Tech Executive Vice President Jeff Olson. “This is an energizing time for all of us at Nor-Tech—knowing that the HPC solutions we are developing for two Nobel projects truly are changing our view of the world.”

    LIGO just announced that their detectors are about to come online after a one-year shutdown for hardware upgrades. In preparation for this, LIGO Consortium member University of Wisconsin-Milwaukee upgraded their clusters with Nor-Tech hardware to assist with the computing demands. At UWM they design, build and maintain computational tools, such as Nor-Tech’s supercomputer, that handle LIGO’s massive amounts of data. Nor-Tech completed the most recent update-including Intel Skylake processors-in 2018. The new Skylake-equipped technology is proving to be almost 10 times faster.

    LIGO was awarded a Nobel Prize in 2017. Prior to this, at a Feb. 11, 2016 national media conference, National Science Foundation (NSF) researchers announced the first direct observation of a gravitational wave. This was a paradigm-shifting achievement in the science community. Subsequent gravitational wave detections have confirmed those results.

    In 2018, the LIGO team announced the first visible detection of a neutrino event. This was made possible, in part, by the powerful HPC technology Nor-Tech has been providing to multiple LIGO Consortium institutions since 2005.

    The first Nor-Tech client to win a Nobel Prize in Physics was the IceCube research team, headquartered at the University of Wisconsin-Madison. IceCube is designed specifically to identify neutrinos from space. It’s a cubic kilometer of ice, laced with photo-detectors, located at a dedicated Antarctic research facility.

    Nor-Tech has been working with several of the world’s leading research institutions involved with the IceCube project for more than 10 years; designing, building, and upgrading HPC technology that made exciting neutrino discoveries possible.

    See the full article here .

    five-ways-keep-your-child-safe-school-shootings

    Please help promote STEM in your local schools.

    Stem Education Coalition

    Founded on December 28, 2006, insideHPC is a blog that distills news and events in the world of HPC and presents them in bite-sized nuggets of helpfulness as a resource for supercomputing professionals. As one reader said, we’re sifting through all the news so you don’t have to!

    If you would like to contact me with suggestions, comments, corrections, errors or new company announcements, please send me an email at rich@insidehpc.com. Or you can send me mail at:

    insideHPC
    2825 NW Upshur
    Suite G
    Portland, OR 97239

    Phone: (503) 877-5048

     
  • richardmitnick 10:52 am on March 25, 2019 Permalink | Reply
    Tags: , , , , , , ExaLearn, insideHPC, , , ,   

    From insideHPC: “ExaLearn Project to bring Machine Learning to Exascale” 

    From insideHPC

    March 24, 2019

    As supercomputers become ever more capable in their march toward exascale levels of performance, scientists can run increasingly detailed and accurate simulations to study problems ranging from cleaner combustion to the nature of the universe. Enter ExaLearn, a new machine learning project supported by DOE’s Exascale Computing Project (ECP), aims to develop new tools to help scientists overcome this challenge by applying machine learning to very large experimental datasets and simulations.

    1
    The first research area for ExaLearn’s surrogate models will be in cosmology to support projects such a the LSST (Large Synoptic Survey Telescope) now under construction in Chile and shown here in an artist’s rendering. (Todd Mason, Mason Productions Inc. / LSST Corporation)

    “The challenge is that these powerful simulations require lots of computer time. That is, they are “computationally expensive,” consuming 10 to 50 million CPU hours for a single simulation. For example, running a 50-million-hour simulation on all 658,784 compute cores on the Cori supercomputer NERSC would take more than three days.

    NERSC

    NERSC Cray Cori II supercomputer at NERSC at LBNL, named after Gerty Cori, the first American woman to win a Nobel Prize in science

    NERSC Hopper Cray XE6 supercomputer


    LBL NERSC Cray XC30 Edison supercomputer


    The Genepool system is a cluster dedicated to the DOE Joint Genome Institute’s computing needs. Denovo is a smaller test system for Genepool that is primarily used by NERSC staff to test new system configurations and software.

    NERSC PDSF


    PDSF is a networked distributed computing cluster designed primarily to meet the detector simulation and data analysis requirements of physics, astrophysics and nuclear science collaborations.

    Future:

    Cray Shasta Perlmutter SC18 AMD Epyc Nvidia pre-exascale supeercomputer

    Running thousands of these simulations, which are needed to explore wide ranges in parameter space, would be intractable.

    One of the areas ExaLearn is focusing on is surrogate models. Surrogate models, often known as emulators, are built to provide rapid approximations of more expensive simulations. This allows a scientist to generate additional simulations more cheaply – running much faster on many fewer processors. To do this, the team will need to run thousands of computationally expensive simulations over a wide parameter space to train the computer to recognize patterns in the simulation data. This then allows the computer to create a computationally cheap model, easily interpolating between the parameters it was initially trained on to fill in the blanks between the results of the more expensive models.

    “Training can also take a long time, but then we expect these models to generate new simulations in just seconds,” said Peter Nugent, deputy director for science engagement in the Computational Research Division at LBNL.

    From Cosmology to Combustion

    Nugent is leading the effort to develop the so-called surrogate models as part of ExaLearn. The first research area will be cosmology, followed by combustion. But the team expects the tools to benefit a wide range of disciplines.

    “Many DOE simulation efforts could benefit from having realistic surrogate models in place of computationally expensive simulations,” ExaLearn Principal Investigator Frank Alexander of Brookhaven National Lab said at the recent ECP Annual Meeting.

    “These can be used to quickly flesh out parameter space, help with real-time decision making and experimental design, and determine the best areas to perform additional simulations.”

    The surrogate models and related simulations will aid in cosmological analyses to reduce systematic uncertainties in observations by telescopes and satellites. Such observations generate massive datasets that are currently limited by systematic uncertainties. Since we only have a single universe to observe, the only way to address these uncertainties is through simulations, so creating cheap but realistic and unbiased simulations greatly speeds up the analysis of these observational datasets. A typical cosmology experiment now requires sub-percent level control of statistical and systematic uncertainties. This then requires the generation of thousands to hundreds of thousands of computationally expensive simulations to beat down the uncertainties.

    These parameters are critical in light of two upcoming programs:

    The Dark Energy Spectroscopic Instrument, or DESI, is an advanced instrument on a telescope located in Arizona that is expected to begin surveying the universe this year.

    LBNL/DESI Dark Energy Spectroscopic Instrument for the Nicholas U. Mayall 4-meter telescope at Kitt Peak National Observatory near Tucson, Ariz, USA


    NOAO/Mayall 4 m telescope at Kitt Peak, Arizona, USA, Altitude 2,120 m (6,960 ft)

    DESI seeks to map the large-scale structure of the universe over an enormous volume and a wide range of look-back times (based on “redshift,” or the shift in the light of distant objects toward redder wavelengths of light). Targeting about 30 million pre-selected galaxies across one-third of the night sky, scientists will use DESI’s redshifts data to construct 3D maps of the universe. There will be about 10 terabytes (TB) of raw data per year transferred from the observatory to NERSC. After running the data through the pipelines at NERSC (using millions of CPU hours), about 100 TB per year of data products will be made available as data releases approximately once a year throughout DESI’s five years of operations.

    The Large Synoptic Survey Telescope, or LSST, is currently being built on a mountaintop in Chile.

    LSST


    LSST Camera, built at SLAC



    LSST telescope, currently under construction on the El Peñón peak at Cerro Pachón Chile, a 2,682-meter-high mountain in Coquimbo Region, in northern Chile, alongside the existing Gemini South and Southern Astrophysical Research Telescopes.


    LSST Data Journey, Illustration by Sandbox Studio, Chicago with Ana Kova

    When completed in 2021, the LSST will take more than 800 panoramic images each night with its 3.2 billion-pixel camera, recording the entire visible sky twice each week. Each patch of sky it images will be visited 1,000 times during the survey, and each of its 30-second observations will be able to detect objects 10 million times fainter than visible with the human eye. A powerful data system will compare new with previous images to detect changes in brightness and position of objects as big as far-distant galaxy clusters and as small as nearby asteroids.

    For these programs, the ExaLearn team will first target large-scale structure simulations of the universe since the field is more developed than others and the scale of the problem size can easily be ramped up to an exascale machine learning challenge.

    As an example of how ExaLearn will advance the field, Nugent said a researcher could run a suite of simulations with the parameters of the universe consisting of 30 percent dark energy and 70 percent dark matter, then a second simulation with 25 percent and 75 percent, respectively. Each of these simulations generates three-dimensional maps of tens of billions of galaxies in the universe and how the cluster and spread apart as time goes by. Using a surrogate model trained on these simulations, the researcher could then quickly run another surrogate model that would generate the output of a simulation in between these values, at 27.5 and 72.5 percent, without needing to run a new, costly simulation — that too would show the evolution of the galaxies in the universe as a function of time. The goal of the ExaLearn software suite is that such results, and their uncertainties and biases, would be a byproduct of the training so that one would know the generated models are consistent with a full simulation.

    Toward this end, Nugent’s team will build on two projects already underway at Berkeley Lab: CosmoFlow and CosmoGAN. CosmoFlow is a deep learning 3D convolutional neural network that can predict cosmological parameters with unprecedented accuracy using the Cori supercomputer at NERSC. CosmoGAN is exploring the use of generative adversarial networks to create cosmological weak lensing convergence maps — maps of the matter density of the universe as would be observed from Earth — at lower computational costs.

    See the full article here .

    five-ways-keep-your-child-safe-school-shootings

    Please help promote STEM in your local schools.

    Stem Education Coalition

    Founded on December 28, 2006, insideHPC is a blog that distills news and events in the world of HPC and presents them in bite-sized nuggets of helpfulness as a resource for supercomputing professionals. As one reader said, we’re sifting through all the news so you don’t have to!

    If you would like to contact me with suggestions, comments, corrections, errors or new company announcements, please send me an email at rich@insidehpc.com. Or you can send me mail at:

    insideHPC
    2825 NW Upshur
    Suite G
    Portland, OR 97239

    Phone: (503) 877-5048

     
c
Compose new post
j
Next post/Next comment
k
Previous post/Previous comment
r
Reply
e
Edit
o
Show/Hide comments
t
Go to top
l
Go to login
h
Show/Hide help
shift + esc
Cancel
%d bloggers like this: