A seldom-seen photo during the blackout inside the Reno convention center at the Student Cluster Competition area.
SCinet’s Year of Epic Power Outages and Road Construction
SCinet, the world’s fastest network built each year for the SC conference by SC volunteers, has a heritage of heroic stories about installing, bringing up, operating and managing networks and systems. Since its inception at SC91, SCinet continuously drives network resources as a platform for high performance computing, storage and analysis, the hallmark of the SC conference. Even in 2007, this story was no different when SC went to Reno, Nevada.
At SC07, SCinet Chair Jackie Kern (SC15’s Conference Chair) led a team of more than 100 volunteers through unexpected power outages and last-minute road de-construction to bring in a whopping 200Gbps of bandwidth (cutting-edge at the time!).
With less than a month before the conference, the well-known “last mile” of fiber-optic infrastructure was needed to reach the convention center and support the network to the show floor.
Jackie Kern, SC07 SCinet Chair
“There was no other way to bring this kind of network connectivity into the convention center, so a trencher was needed to complete some last-minute road work—literally cutting the roadway outside the convention center to lay fiber into the facility,” explained Kern. This was patched with asphalt and served as a reminder that nothing stands in the way of SC’s SCinet network, which today is still a lasting connection into the Reno-Sparks Convention Center.
In addition to SCinet’s dedication to bringing in the technology to make demos and applications at SC possible, during the SC07 Exhibit show Kern said “a moment of complete silence occurred when all electronic devices suddenly came to a halt.” All was quiet as the convention center experienced three “power bumps,” overrunning all the circuits to the convention center with excessive electric currents. When power was restored moments later, hundreds of devices attempted to come back online, once again overriding the facility’s power infrastructure. With coordinated and staged power up, the SCinet team worked relentlessly to restore service to all exhibitors ASAP. The restoration of the SCinet network was tedious, but all services were returned to normal operations within a few hours.
The SCinet team is always on their toes, ready to respond to a variety of inevitable and unforeseeable hiccups. That said, all volunteers would agree these amazing, priceless experiences contribute to learning opportunities available only by building the fastest network in the world that is planned for one year, set up in one month, operated for one week, and torn down in 24 hours!
Student volunteers are a key component of each SC conference.
The SC15 Student Volunteers program is still accepting applications, but the deadline to submit is Monday June 1. Student Volunteers play a key role in supporting the conference, and this year the program is expanding both in the number of volunteer participants and the quality of the student programs in which the volunteers will have access.
Undergraduate and graduate student volunteers help with the administration of the conference, participate in student-oriented activities, attend technical talks by famous researchers and industry leaders, explore the exhibits and develop lasting peer connections. Read more by clicking here.
Austin, TX – May 20, 2015 –The Supercomputing Conference (SC15) Test of Time Award Committee has recognized “The NAS Parallel Benchmarks - Summary and Preliminary Results” written by D. Bailey, E. Barszcz, J. Barton, D. Browning, R. Carter, L. Dagum, R. Fatoohi, P. Frederickson, T. Lasinski, R. Schreiber, H. Simon, V. Venkatakrishnan, and S. Weeratunga from the SC91 conference as the SC15 Test of Time Award (ToTA) paper for this year.
The ToTA recognizes an outstanding paper that has appeared at the SC conference and has deeply influenced the HPC discipline. It is a mark of historical impact and recognition that the paper has changed HPC trends, and will be presented at the SC15 conference in a non-plenary session and the authors will be asked to give a presentation on the work in Austin, TX in November 2015. To view this specific paper, click here.
“The paper and benchmark captures specifications and implementations of an important set of representative scientific codes,” said Jack Dongarra, SC15 Test of Time Award Chair from University of Tennessee, Knoxville. “The work is still actively used, and has inspired numerous sets of benchmarking codes that continue to drive research and development innovation.”
In 1991, this team of computer scientists from the Numerical Aerodynamic Simulation Program—predecessor to the NASA Advanced Supercomputing (NAS) facility at Ames Research Center—unveiled the NAS Parallel Benchmarks (NPB), developed in response to the U.S. space agency’s increasing involvement with massively-parallel architectures and the need for a more rational procedure to select supercomputers to support agency missions. Then, existing benchmarks were usually specialized for vector computers, with shortfalls including parallelism-impeding tuning restrictions and insufficient problem sizes, making them inappropriate for highly-parallel systems.
The NPBs mimic computation and data movement of large-scale computational fluid dynamics (CFD) applications, and provide objective evaluation of parallel HPC architectures. The original NPBs featured “pencil-and-paper” specifications, which bypassed many difficulties associated with standard benchmarking methods for sequential or vector systems. The principal focus was in computational aerophysics, although most of these benchmarks have broader relevance for many real-world scientific computing applications.
The NPBs quickly became an industry standard and have since been implemented in many modern programming paradigms. Since 1991, research areas influenced by the NPBs have broadened to include network design, programming languages, compilers, and tools. Google Scholar yields over 27,000 results for the NBPs, with about 7,400 citations since 2014.
Today’s version is alive and well, and continues to significantly influence NASA projects. It is used around the world by national labs, universities, and computer vendors to evaluate sustained performance of highly-parallel supercomputers and the capability of parallel compilers and tools.
AUSTIN, Texas–Each year, the global supercomputing community honors a handful of the leading contributors to the field with the presentation of the IEEE Seymour Cray Computer Science and Engineering Award, the IEEE Sidney Fernbach Memorial Award and the ACM-IEEE Ken Kennedy Award.
Nominations for these awards to be presented at SC15 in Austin are now open and the submission deadline is Wednesday, July 1, 2015. Recipients of this year’s awards will give special presentations during SC15, to be held Nov. 15-20 at the Austin Convention Center. Here are descriptions of the three awards:
IEEE Seymour Cray Computer Science and Engineering Award
Seymour Cray
Established in 1997, the IEEE Computer Society Seymour Cray Computer Engineering Award recognizes innovative contributions to high performance computing systems that best exemplify the creative spirit demonstrated by Seymour Cray. Previous winners have been recognized for design, engineering and intellectual leadership in creating innovative and successful HPC systems. One of HPC’s most prestigious awards, winners receive an illuminated certificate and a $10,000 honorarium at a special awards session during the conference. Learn more by clicking here.
Sidney Fernbach
IEEE Sidney Fernbach Memorial Award
The IEEE Computer Society Sidney Fernbach Award was established in 1992 in honor of Sidney Fernbach, one of the pioneers in the development and application of high performance computers for solving large computational problems. Nominations that recognize creation of widely-used and innovative software packages, application software and tools are especially solicited. The Fernbach award winner receives a certificate and $2,000. Learn more by clicking here.
ACM-IEEE CS Ken Kennedy Award
Ken Kennedy
The ACM/IEEE Ken Kennedy Award, established in 2009, is presented for outstanding contributions to programmability or productivity in computing, together with significant community service or mentoring contributions. The award was established in memory of Ken Kennedy, the founder of Rice University's nationally ranked computer science program and one of the world's foremost experts on high performance computing. Awardees receive a certificate and a $5,000 honorarium. Learn more by clicking here.
Questions on the awards may be directed to: awards@info.supercomputing.org.
An enlightening
video series launched by the SC conference steering committee in 2013
aims to illustrate how high performance computing is impacting everyday
life – from manufacturing to storm prediction to the making of Hollywood
blockbusters. The latest in the series is a short video highlighting
the innovative work being done at the University of Utah’s Scientific
Computing and Imaging Institute in regards to helping Parkinson’s
patients lead more normal lives through deep brain stimulation (DBS). The Institute helps doctors pinpoint brain stimulation sites
that relieve tremors in Parkinson’s patients and drastically improve
quality of life.
Here's the short video and accompanying article with specific details.
Although it may sound like something straight from a scene in a science fiction film, new surgery techniques that place a set of wires under the skull to transmit electrical signals to different areas of the brain has gotten even more effective with the help of computers. It's called deep brain stimulation, or DBS. And if the idea of it seems a bit wince-inducing or scary, then understanding the power of what it can do - quiet the tremors associated with Parkinson's disease and other brain disorders - will likely wash away any patient's fears and it is being enhanced via high performance computing and visualization.
FDA-approved in 1997, DBS involves the surgical implantation of a neurostimulator device - a sort of brain pacemaker - into the chest along with a set of wires, known as leads, which together send electrical impulses to various parts of the brain in order to activate brain circuits. This can alleviate the symptoms associated with movement disorders, most commonly Parkinson's, essential tremor, and dystonia.
Computational models of DBS are created by combining medical imaging,
bioelectric field models and populations of multicompartmental neuron
models. Butson & McIntyre, Brain Stim, 2008.
"We know DBS does not change the progression of Parkinson's disease, but it does improve people's quality of life after they have it," says Paul House, MD, a neurosurgeon and the surgical director of University of Utah Health Care's Movement Disorders Program who himself has done more than 400 of the surgeries at University of Utah Health Care (UUHC). According to Dr. House, they get a lot more years of improved quality of life. Nationally, UUHC ranks in the top 15 for number of annual DBS procedures with some of the best patient outcomes.
Recently the program hired Christopher Butson, PhD, as its director of neuromodulation research. An expert in neurostimulation devices and neuromodulation therapy, Dr. Butson came to Utah from the Biotechnology and Bioengineering Center at the Medical College of Wisconsin in Milwaukee. One of Dr. Butson’s goals is to use neurostimulation therapy to improve the lives of patients with a range of neurological and psychiatric disorders.
Another key component in Dr. Butson's work: data. UUHC has an extensive high performance computing infrastructure that is the raw material for much of his work. Much of those rich details were not - but now will be - mined for contributions to new discovery and innovation.
Computational models can make detailed predictions about the
effects of stimulation in individual subjects based on biophysical
tissue properties derived from diffusion tensor imaging. In the bottom
right image, an area of DBS-induced activation are shown as it impinges
on surrounding nuclei. Butson et al, NeuroImage, 2007.
"The one thing we aspire to do in my lab is to combine computational models with patient data to create insights that would be difficult to achieve otherwise," he says. One of Dr. Butson's most promising ideas: an iPad based interactive computer program that creates a three-dimensional picture of the brain. This provides clinicians with a better look at a patient's brain and helps them make better predictions about DBS lead placements, stimulation settings and effects.
It also saves time. Clinicians who used the program in a small pilot study reduced device programming time by 99 percent. The program, which allows for simulated stimulation of the brain's circuits, may eventually help enhance the precision of surgical targets and DBS treatment for a multitude of disorders and diseases. Dr. Butson and Co-PI Dr. Michael Okun at the University of Florida were recently awarded a grant from the National Institutes of Health to conduct a clinical trial of this system in a much larger cohort of patients. It's something that until recently couldn't really be done, but has been made possible by advances in medical imaging technology - another area of research that UUHC helped advance.
"Where the field is going is toward circuit-based therapy... and we hope that 3D modeling will allow us to see this in a totally different way," Dr. Butson says. "So integrating the best information we can get from all different types of imaging and incorporating that into predictive models... this is the bleeding edge of DBS research."
More About SC15:
SC15, sponsored by the ACM (Association for Computing Machinery) and the IEEE Computer Society, offers a complete technical program, programs for students and educators in HPC, and an exhibition that together showcase the many ways high performance computing, networking, storage and analysis lead to advances in scientific discovery, research, education and commerce. This premier international conference includes a globally attended technical program, workshops, tutorials, a world class exhibit area, demonstrations and opportunities for hands-on learning.
More About the University of Utah Scientific Computing and Imaging (SCI) Institute:
The SCI Institute is an interdisciplinary research institute consisting of more than 200 faculty, staff, and students. It has established itself as an internationally recognized leader in visualization, scientific computing, and image analysis applied to a broad range of important application. A particular hallmark of SCI Institute research is the development of innovative and robust software packages, including the SCIRun scientific problem solving environment, Seg3D, ImageVis3D, FEBio, Shapeworks, and FluoRender. All these packages are broadly available to the scientific community under open source licensing and supported by web pages, documentation, and users groups.
SCinet, the world’s fastest network built each year for the SC conference by SC volunteers, has a heritage of heroic stories about installing, bringing up, operating and managing networks and systems. Since its inception at SC91, SCinet continuously drives network resources as a platform for high performance computing, storage and analysis, the hallmark of the SC conference. Even in 2007, this story was no different when SC went to Reno, Nevada.
