At insideHPC, we are very pleased to bring you live streaming keynotes from the GPU Technology Conference this week in San Jose.
In this video, Erez Lieberman Aiden from the Baylor College of Medicine presents a keynote talk entitled: Parallel Processing of the Genomes, by the Genomes and for the Genomes.
The human genome is a sequence of 3 billion chemical letters inscribed in a molecule called DNA. Famously, short stretches (~10 letters, or “base pairs”) of DNA fold into a double helix. But what about longer pieces? How does a 2 meter long macromolecule, the genome, fold up inside a 6 micrometer wide nucleus? And, once packed, how does the information contained in this ultra-dense structure remain accessible to the cell? This talk will discuss how the human genome folds in three dimensions, a folding enables the cell to access and process massive quantities of information in parallel. To probe how genomes fold, we developed Hi-C, together with collaborators at the Broad Institute and UMass Medical School. Hi-C couples proximity-dependent DNA ligation and massively parallel sequencing. To analyze our data and reconstruct the underlying folds, we, too must engage in massively parallel computation. I will describe how we use NVIDIA’s CUDA technology to analyze our results and simulate the physical processes of genome folding and unfolding.
In this special guest feature, Doug Black from The Exascale Report writes that, while the idea of Grand Challenges is not new, the need for powerful computational tools to solve these global issues remains unchanged.
Flash back to 1992. Do you remember the ‘Blue Book’ and the HPCC program? If this is your first exposure to the ‘Grand Challenges’ you may find this quite interesting. On November 7, 2012, senior representatives of the DOE labs sent a letter to Secretary of Energy, Steven Chu to report on a Grand Challenges Workshop on Advanced Computing for Energy Innovation held in late July – early August 2012.
While the workshop recommendations focused on what it called Technical, Structural and Incentive ‘Grand Challenges’, one of its final recommendations was to establish an Advanced Computing for Energy (ACE) program within the Department of Energy. When I read this letter, I had an intense sense of déjà vu – one of those ‘here we go again’ feelings. But in a good way.
For a moment, it felt like 1992 all over again, a year of unusually high energy and high promise in the HPC community. It’s the year we really sank our teeth into the teraFLOPS challenge. It seemed the entire community rallied in support of what the first President Bush’s science advisor, Alan Bromley, labeled the Grand Challenges – referring to high performance computing and communications. Those Grand Challenges were the challenges of science.
It was the beginning of a period of powerful government and private industry collaboration referred to as the HPCC program. I pulled this quote from the program’s overview documentation: The HPCC Program is driven by the recognition that unprecedented computational power and capability is needed to investigate and understand a wide range of scientific and engineering “grand challenge” problems.
The program’s famous “Blue Book” also made this point:
The HPCC Program is the result of several years of effort on the part of senior government, industry, and academic scientists and managers to design a research agenda to extend U.S. leadership in high performance computing and networking technologies.
So, in many ways, nothing has really changed. Again, I mean this in a good way. The 2012 appeal to address the world’s ‘Grand Challenges’ is eerily similar to what we addressed 20 years ago. HPC is an ever widening circle that keeps coming around. Twenty years ago, the Grand Challenges included climate prediction and genome mapping. Today, the great need is energy innovation and saving the environment. Tomorrow, it may be food. This is HPC and that’s how HPC works, tackling as ever the need for funding and the need for urgency to apply extreme computational resources on the greatest scientific challenges of our time.
Today Nvidia announced that growing ranks of Python users can now take full advantage of GPU acceleration for HPC and Big Data analytics applications by using the CUDA parallel programming model. As a popular, easy-to-use language, Python enables users to write high-level software code that captures their algorithmic ideas without delving deep into programming details. Python’s extensive libraries and advanced features make it ideal for a broad range of HPC science, engineering and big data analytics applications.
Our research group typically prototypes and iterates new ideas and algorithms in Python and then rewrites the algorithm in C or C++ once the algorithm is proven effective,” said Vijay Pande, professor of Chemistry and of Structural Biology and Computer Science at Stanford University. “CUDA support in Python enables us to write performance code while maintaining the productivity offered by Python.”
Support for CUDA parallel programming comes from NumbaPro, a Python compiler in the new Anaconda Accelerate product from Continuum Analytics. This support was made possible by Nvidia’s contribution of the CUDA compiler source code into the core and parallel thread execution backend of LLVM, a widely used open source compiler infrastructure. Read the Full Story.
The GPU Technology Conference kicks off next week in San Jose with a focus on scientific computing. The conference has grown considerably over the years, so I caught up with NVIDIA’s Andy Walsh to learn more. Walsh currently serves as Director of Marketing for their Accelerated Computing Business.
insideHPC: What will be new and exciting this year for the HPC crowd at GTC?
Andy Walsh: We’ve come a long way since the first GTC. In 2009, we created GTC to engage a diverse group of developers, scientists, and researcher using CUDA GPUs. The response then was much greater than we could imagine.
GTC has become the most important event showcasing breakthroughs in science and industry, thousands of the brightest minds will gather at GTC to meet, network and share ideas. This year’s conference will feature 400+ sessions and attendees from over 40 countries. What’s new this year is we expanded the scope to include tracks in manufacturing and design, media and entertainment, cloud computing, game and mobile development, and more. Attendees also will experience interesting things up and down the concourse from test driving state of the art cars to over 150 research posters.
insideHPC: How big are the exhibits this year?
Andy Walsh: GTC 2013 is expected to have our best exhibit area ever, with the largest numbers of companies exhibiting. This year, over 100 of the most important companies in the industry will be showcasing cutting edge offerings from HPC to mobile technology. We are also featuring companies that have joined us for the very first time, including Cisco, Citrix, VMware, IGI, Microsoft, Ingram Micro, Quantel, and Acer. Lunches and cocktail receptions are held in the exhibit hall, making the GTC exhibits area a vibrant place to network and learn about GPU solutions.
insideHPC: I’m particularly excited about the Emerging Companies Summit. Can you tell us more about that?
Andy Walsh: The Emerging Companies Summit (ECS) provides an opportunity for startups to showcase how they are using GPUs to transform industries and create new ones. This year, ECS will feature 16 companies from around the world advancing diverse fields, as well as a “CEO on Stage” event – where executives will present their companies to a panel of investors, analysts and tech leaders who will challenge the presenters with questions and provide insightful feedback. Five top startups will be recognized for their innovation in a competition with more than $75,000 in prizes.
insideHPC: When does it all begin next week?
Andy Walsh: GTC 2013 starts on Monday, March 18, with a full-day of preconference tutorials on a variety of topics including GPU programming languages and libraries to application optimization, visualization, video processing, and ray tracing.
On Tuesday, March 19, NVIDIA CEO and Co-Founder Jen-Hsun Huang will officially kick off a week of GTC sessions with the delivery of the opening keynote.
Today Fujitsu announced the launch of operations of the purpose-built Atacama Compact Array (ACA) Correlator supercomputer system, which will process images from the Atacama Large Millimeter/submillimeter Array (ALMA) project, a Chile-based radio telescope featuring unprecedented sensitivity and resolution.
With the observations from ALMA, we hope to gain insights into such mysteries as how galaxies have formed and evolved, how planetary systems orbiting around a Sun-like star are formed, and whether the origin of life is to be found in the universe. The data processing performed by the ACA Correlator system is essential for these types of radio astronomy research. I am confident that ALMA will open new horizons for astronomy.”
Fujitsu and the National Astronomical Observatory of Japan (NAOJ) worked together to develop the ACA Correlator, a purpose-built supercomputer responsible for processing data from the Atacama Compact Array, which can make high sensitivity observations.
Set at 5,000 meters above sea level in the Chilean Andes, ALMA is a massive radio telescope developed through a partnership among East Asia (led by NAOJ), North America and Europe. The telescope is capable of producing astronomical radio wave images with the world’s highest resolution. The facility consists of 66 antennas arranged in a 18.5 km-diameter array, equivalent to the span of the Yamanote railway loop encircling the central part of Tokyo, and by processing millimeter/submillimeter wave(1) signals from each antenna, it is possible for the antennas to act as a single, giant telescope that can generate radio wave images with the same resolution as those produced by a massive 18.5 km-diameter parabolic antenna. This makes it possible to see the dark regions of the universe that cannot be observed at optical wavelengths, such as galaxies that were formed shortly after the beginning of the universe, the birth of stars, planetary systems like our solar system, and matter related to the origin of life, such as of organic molecules.
A ceremony was held in Chile to commemorate the inauguration of ALMA on March 13. Read the Full Story.
Information will be the most valuable resource in the 21st century. Operating on large volumes of diverse data sources to get the right actionable insights at the right time presents new challenges and opportunities for system design. Addressing these opportunities requires a rethinking of future server and data center design—with a datacentric focus across both hardware and software. Here, we’ve presented a brief introduction to some recent research activities in this exciting emerging area, with a specific focus on system architecture and systems software.
The idea that we use only 10 percent or less of our brain is one of those persistent myths that stubbornly refuses to go away. In reality, that bit of gelatinous grey matter between our ears has been extensively mapped and it appears that most of it has a function. So the notion that if only we could harness that underused 90 percent of our brain we could develop superhuman mental abilities, needs to be relegated to the urban legend scrap heap.
But what about the idea that the world of Big Data is in even worse shape – that only one percent of what’s actually out there is being analyzed?
That’s the contention of Gurjeet Singh, a guest contributor to a recent issue of GigaOM.
Singh is the co-founder and CEO of Ayasdi, a discovery platform built on topological data analysis technology. He says that to make use of the untouched 99 percent of available data – some one quintillion bytes that are collected every day (or, according to IBM, a daily harvest of 2.5 quintillion bytes) – we have to “fundamentally change the way we gain knowledge from data.”
Singh calls on the next wave of Big Data solutions to: empower domain experts such as biologists, geologists, etc. – not just the data scientists; accelerate the pace of discovery so we can get insights faster; create new levels of machine intelligence – human thought is too slow for many Big Data operations; and design systems that are able to efficiently analyze both structured and unstructured data, especially the latter in all its myriad forms.
When it comes to the evolution of big data, we’ve only begun to scratch the surface,” Singh writes. “It stands to reason that if we continue to analyze 1 percent of data, then we’ll only tap into 1 percent of its potential. If we’re able to analyze the other 99 percent, then think about all of the ways that we can change the world. We can accelerate economic growth, cure cancer and other diseases, reduce the risk of terrorist attacks, and many other big ticket challenges that we’re faced with. That’s something that we can all rally around.”
In the supercomputing world, we often think of things in tens. Over at TACC, Aaron Dubrow has posted a series of interviews with ten of the top HPC minds in Texas. With a focus what on what terascale, petascale, and exascale means to them and their field, the interviews cover a broad base of application user space.
Recently, a mouse with diminished interferon (proteins made and released by host cells in response to the presence of pathogens such as viruses, bacteria, parasites or tumor cells) was identified in our laboratory, said Bruce Beutler, Regental Professor and Director, Center for Genetics of Host Defense, UT Southwestern Medical Center. “Because we had sequenced the genomic DNA of its grandfather, we knew that the mouse likely had a mutation in a gene coding for the protein kinase TBK1. Without any mapping, we determined the exact cause of the phenotype. In former times, before it was possible to routinely sequence the genome of these mice, we would have spent months and thousands of dollars arriving at this same conclusion. This is a hint of the speed and dexterity that advanced computing can provide. It permits us to exclude obvious causes of phenotype and concentrate on what is new. Our research would be impossible without enormous computational resources.
Since 2012, PRACE has been offering European companies an industrial R&D service, based on a set of complementary high-level services including information and networking, training, access to leading HPC resources and expertise, and code-enabling of open-source applications. Now this documented project offer has been recognised as an effort in catalysing European industrial competitiveness.
The FP7 Success Story Competition highlights the three best success stories from the FP7 Capacities funding programme in e-Infrastructures. Project success stories in the competitive industry category show how the project made Europe a more attractive location to invest in research and innovation, by promoting activities where businesses set the agenda. The project is aimed at helping innovative SMEs to grow into world-leading companies.
Since the establishment of the PRACE Open R&D industrial offer in January 2012, PRACE has been able to attract more than 10 European companies; large companies as well as SMEs (Small and Medium Enterprises) for using its HPC facilities as well as the other high-value services.
This award in ‘competitive industries’ will foster our motivation to work on engaging industrial users on the PRACE research infrastructure in order to boost European competitiveness,” said Stephane Requena, author of the project paper and member of the board of directors of PRACE. “PRACE is working on increasing the use of a leading European infrastructure by all academic and industrial communities and is catalysing technological transfer between academia and industry through open innovation projects. In that sense we are working in the field of the FP7 funded PRACE-3IP implementation project on a tailored evangelisation programme called SHAPE (SME HPC Adoption Programme in Europe) which aims to help SMEs to co-design and demonstrate a concrete industrial project on PRACE facilities.”
The HPC Midlands supercomputing facility will host their launch event in the U.K. on March 20. As a provider of state-of-the-art e-infrastructure for research and industry, HPC Midlands features a 3,000 core supercomputer combined with HPC expertise from Loughborough University and the University of Leicester.
Since establishing HPC Midlands with the financial backing of the Engineering and Physical Sciences Research Council, we have worked closely with academic colleagues and a range of industrial partners to refine the service to ensure that it meets business as well as academic needs,” said Dr Steven Kenny, Director of HPC Midlands. “Now we are ready to invite small and large businesses with specialist computing requirements to come along and see how they can benefit from this world-class facility.”
The launch event will give delegates the chance to meet the team behind HPC Midlands and explore opportunities for collaboration. Case study presentations will showcase how company’s like Tata Steel, E.ON, and Rolls Royce already benefit from working closely with HPC Midlands. Read the Full Story.
Fans of the old Sun Microsystems may be wondering how the server business is doing at Oracle some three years after the acquisition. Over at GigaOm, Barb Darrow writes that Oracle’s gamble on hardware just isn’t paying off.
Here’s the problem, since it entered the hardware business, Oracle hasn’t sold enough engineered systems to make up for lost sales of lower-end machines, according to third-party researchers. Its hardware revenue and unit share is headed south. For the fourth calendar quarter of 2012, Oracle server revenue was down 18 percent year over year according to both Gartner and IDC. Meanwhile, as GigaOM’s Jordan Novet reported last week, the “other” server vendors — companies like Quanta and Wistron – saw their aggregate revenue rise nearly 22 percent in the fourth quarter compared to the year-ago period.
There is no doubt that the Sequester is putting extreme pressure on critical R&D in this country. As members of the HPC community, how can we help to ensure that we continue to fund important supercomputing research?
In this video from February 15, 2013, the House Subcommittee on Research holds a hearing on Applications for Information Technology Research & Development.
Witnesses:
Dr. Kelly Gaither, Director, Visualization Lab, Texas Advanced Computing Center, University of Texas, Austin
Dr. Ed Lazowska, Bill and Melinda Gates Chair in Computer Science and Engineering, University of Washington
The topic of this afternoon’s hearing, Applications for Information Research & Development, is important to our national security, global competitiveness and technological innovation. This hearing will provide us with examples of practical applications and the benefits of Federal investment in networking and information technology R&D. The Networking and Information Technology Research and Development program, or NITRD, was originally authorized in 1991 in the High Performance and Computing Act. It coordinates the networking and information R&D efforts of 15 Federal member agencies, including DHS, NASA, NIH, EPA and the Department of Energy. The program is the main R&D investment portfolio of member agencies in networking, computing, software, cyber security and related information technologies totaling over $3.7 billion in FY2013.
On behalf of insideHPC, I just want to say, “Great job, Kelly!” Read the Full Story.
This is a test page that uses the same script as our ad server. Scroll down to the bottom to see the ads.
Today NetApp announced the new NetApp E5500 storage system for Big Data and HPC applications. Derived from Engenio technology acquired from LSI, the seventh-generation E-Series is a high-density platform with record per-spindle throughput.
The SGI InfiniteStorage 5600, which is an OEM version of the NetApp E5500, has produced a new SPC-2 result confirming the performance and cost efficiency of the new E5500; it showcases the performance possibilities the E5500 unlocks for HPC and big data organizations. The audited, peer-reviewed SPC-2 result demonstrates the highest throughput per spindle by more than 2.5 times over the nearest non-NetApp published result.
Read the Full Story.
In related news, SGI rolled out the InfiniteStorage 5600 storage system today, noting that the flexibility of the platform enables users to push the limit without breaking their budget.
inShare
Comment on this story by Ralph on Mar 6
Advertisement
Accelerate x64 Applications on GPUs Using the Most Trusted HPC Compilers: The Portland Group.
The Unknowns that come with Exascale
Posted in Exascale, HPC
Over at Reed’s Ruminations, Dan Reed writes that the road to Exascale will come complete with the unknowns that make up any change that involves orders of magnitude.
All currently envisioned exascale systems would require parallelism at unprecedented scale, and barring new, energy efficient memory technologies; they would be memory starved relative to current systems, even under a 20 MW system design point; and multilevel fault tolerance would be required to achieve acceptable systemic mean time to failure (MTBF). Extraordinary parallelism, unprecedented data locality and adaptive resilience: these are daunting architecture, system software and application challenges for exascale computing.
Read the Full Story.
inShare
Comment on this story by Ralph on Mar 6
Video: Supercomputer Modelling of a Complete Human Viral Pathogen
Posted in HPC, Video, Visualization
In this video, researchers perform a reconstruction and simulation of the poliovirus using the BlueGene/Q supercomputer at the Victorian Life Sciences Computation Initiative. The poliovirus model is being used as a basis for understanding antiviral drugs, virus infection and helps us to learn how to model related viruses such as Enterovirus 71.
Read the Full Story.
inShare
Comment on this story by Rich Brueckner on Mar 6
Sponsored Post: OpenFabrics Software Events Coming to Monterey in April
Posted in Events, HPC, HPC Hardware, HPC Software, Network, Open Fabrics, Open Fabrics Workshop, Sponsored Post
Be a part of the OpenFabrics Alliance’s OpenFabrics Software events in Monterey, CA.
The OFA User Workshop, April 18-19, provides opportunities to share experiences and learn from a community of OFS users.
The International Developer’s Workshop, April 21-24, will focus on the development and improvement of OFS as well as major developments in RDMA, etc. Agenda and more information is available on OpenFabrics.org.
Registration for the two events is now open. For more details, check out the OFA Newsletter.
inShare
Comment on this story by Rich Brueckner on Mar 5
Video: Enterprise HPC in the Cloud – Fortune 500 Use Cases
Posted in Cloud HPC, Events, Video
Over at the Compute Cycles Blog, Jason Stowe from Cycle Computing writes that the recent AWS: reInvent conference featured a series of talks on Enterprise Use Cases of HPC in the Cloud. In this video, Stowe introduces Cloud HPC talks from Johnson & Johnson, Pacific Life, Hartford Insurance, and Life Technologies.
Read the Full Story.
inShare
Comment on this story by Rich Brueckner on Mar 5
Benchmarking Intel Xeon Phi vs. Sandy Bridge
Posted in Co-processors, Compute, HPC, HPC Hardware
Intel has been careful to label the Xeon Phi as a coprocessor, something that always pairs with a Xeon CPU. But how does their performance compare on real applications? Over at the Xcelerit Blog, Paul Sutton benchmarks both devices using an optimized parallel version of the Monte-Carlo LIBOR swaption portfolio pricer.
It is executed once on the host CPUs (the Sandy Bridge processors), and again on the Xeon Phi co-processor in offload mode. The execution time of the full application is measured, including data transfers, random number generation, and reduction. All these steps are running on the target processor.
As we can see, from about 100K paths onwards, the Intel Xeon Phi becomes faster than the Sandy Bridge processors, reaching nearly 3x at 1M paths. With lower numbers of paths, the Sandy Bridge outperforms the Phi. This can be explained by the added data transfers and the comparably low level of parallelism for a low number of paths (considering both vectorization and multi-threading). The setup time for the random number generator also becomes more dominant on the Xeon Phi when there is relatively little computation performed.
Read the Full Story.
inShare
Comment on this story by Ralph on Mar 5
Podcast: Student Cluster Competition Comes to South Africa, Winners off to ISC’13
Posted in Events, HPC, ISC13, Podcast, Rich Report
In this podcast, Happy Sithole from CHPC and Gilad Shainer from Mellanox discuss the first-ever Student Cluster Competition in South Africa, which wrapped up in December. The winning team will be traveling to compete in the ISC’13 Student Cluster Competition in Leipzig this June.
Download the MP3 or Subscribe on iTunes.
In related news, you can now follow all the major worldwide Student Cluster Competitions at their new home site.
inShare
Comment on this story by Rich Brueckner on Mar 5
1 Terabyte/sec File Systems Enable Big Fast Data
Posted in HPC, HPC Hardware, Storage
Over at the new Cray Blog, VP of Storage Barry Bolding writes that Big Data at the company is synonymous with very fast data access. As evidenced by the Lustre-powered Terabyte/sec file system on Blue Waters, the time for Big Fast Data has come.
Success was achieved on Oct. 7, 2012 when Cray and NCSA measured 1.037 TB/sec of performance to a single Lustre filesystem and 1.137 TB/sec of simultaneously to all three Lustre filesystems. To put this achievement into mass media perspective, 1TB/sec is equivalent to downloading about 125 copies of Dead Space 3 per second, or about 250,000 songs through Spotify per second. This is a milestone for parallel filesystems, a milestone for Lustre, a milestone for Cray and a milestone for our customer at NCSA. Best of all this the technology is being put to good use by a cadre of National Science Foundation users in solving some of the most challenging scientific problems facing our world today.
Over at Reed’s Ruminations, Dan Reed writes that the road to Exascale will come complete with the unknowns that make up any change that involves orders of magnitude.
All currently envisioned exascale systems would require parallelism at unprecedented scale, and barring new, energy efficient memory technologies; they would be memory starved relative to current systems, even under a 20 MW system design point; and multilevel fault tolerance would be required to achieve acceptable systemic mean time to failure (MTBF). Extraordinary parallelism, unprecedented data locality and adaptive resilience: these are daunting architecture, system software and application challenges for exascale computing.
The idea that we use only 10 percent or less of our brain is one of those 
Over at 
