{"62674":{"#nid":"62674","#data":{"type":"news","title":"Georgia Tech Keeps High Performance Computing Sights Set on Exascale at SC10","body":[{"value":"\u003Cp\u003EThe road to exascale computing is a long one, but the Georgia \nInstitute of Technology, a new leader in high-performance computing \nresearch and education, continues to win new awards and attract new \ntalent to drive technology innovation. From algorithms to architectures \nand applications, Georgia Tech\u0027s researchers are collaborating with top \ncompanies, national labs and defense organizations to solve the complex \nchallenges of tomorrow\u0027s supercomputing systems. Ongoing projects and \nnew research initiatives spanning several Georgia Tech disciplines \ndirectly addressing core HPC issues such as sustainability, reliability \nand massive data computation will be on display November 13-19, 2010 at \nSC10 in New Orleans, LA. \u003C\/p\u003E\n\t\u003Cp\u003ELed by Jeffrey Vetter, joint professor of computational science and \nengineering at Georgia Tech and Oak Ridge National Laboratory, Keeneland\n is an NSF-funded project to deploy a high-performance heterogeneous \ncomputing system consisting of HP servers integrated with NVIDIA Tesla \nGPUs. Entering its second-year, the project will deploy its initial \ndelivery system \u2013 the first of two experimental systems \u2013 this month. \nDuring the initial performance runs, the Keeneland system was clocked at\n running 64 teraflops per second, placing it well within the top 100 \nsystems in the world on the most recent TOP500 list of supercomputers \n(June 2010). Given the system\u0027s excellent energy efficiency of \napproximately 650 megaflops per second per watt on the TOP500 Linpack, \nthe team is hoping to secure a strong position on the Green500 list of \nthe most energy efficient supercomputers in the world. Keeneland is \nsupported by a $12 million grant from NSF\u0027s Track 2D program, a \nfive-year activity designed to fund the deployment and operation of two \ninnovative  computing systems, with an overarching goal of preparing the\n open computational science community for emerging architectures that \nhave high performance and are energy efficiency.\u003C\/p\u003E\n\t\u003Cp\u003E\u0022Heterogeneous computing will play an important role in the future \nof high performance computing due to the new challenges of extreme \nparallelism and energy efficiency,\u0022 said Vetter. \u0022The Keeneland \npartnership is providing hardware and software resources, training, and \nexpertise to the computational science community at a critical time in \nthis transition to new computing architectures.\u0022\u003C\/p\u003E\n\t\u003Cp\u003EA Georgia Tech team led by George Biros is a Gordon Bell Prize \nfinalist at SC10 for their work demonstrating the simulation of blood \nflow using heterogeneous architectures and programming models at the \npetascale using CPU and hybrid CPU-GPU platforms, including the new \nNVIDIA Fermi architecture and 200,000 cores of ORNL\u0027s Jaguar system.\u003C\/p\u003E\n\t\u003Cp\u003EReliable and sustainable computing are core aspects of DARPA\u0027s \nrecently announced Ubiquitous High Performance Computing (UHPC) program,\n a $100 million initiative to build future systems that dramatically \nreduce power consumption while delivering a thousand-fold increase in \nprocessing capabilities. Georgia Tech researchers are supporting several\n components of the NVIDIA-led UHPC team, ECHELON, while the Georgia Tech\n Research Institute (GTRI) will lead a fifth group, CHASM, that will \ndevelop applications, benchmarking and metrics to drive UHPC system \ndesign considerations and support performance analysis of the developing\n system designs.\u003C\/p\u003E\n\t\u003Cp\u003E\u0022The key to solving the energy requirement roadblock to future \nsystems is massive parallelism, which requires an entirely new way of \nthinking about today\u0027s algorithms and architectures,\u0022 said Dan Campbell,\n senior researcher at GTRI and a co-PI of CHASM. \u003C\/p\u003E\n\t\u003Cp\u003E\u0022UHPC provides an opportunity for anticipated application challenges\n to influence the high-end system designs, in ways that are outside the \ntraditional planning of industrial roadmaps in high performance \ncomputing,\u0022 said David Bader, professor of Computational Science \u0026amp; \nEngineering at Georgia Tech and Applications Lead for ECHELON. \u003C\/p\u003E\n\t\u003Cp\u003EGeorgia Tech was also named an NVIDIA CUDA Center of Excellence in \nAugust 2010, further empowering the Institute to conduct game changing \nresearch and increase the computing power available to scientists and \nengineers through massively parallel computing.\u003C\/p\u003E\n\t\u003Cp\u003EWhile computing systems one thousand times faster than current \npetascale levels is still 10 years away, massive amounts of data are \ncurrently being generated every day in health care, computational \nbiology, homeland security, commerce, social media and many other \nindustries. Georgia Tech is attacking the massive data analytics \nchallenge. The Georgia Tech-led Foundations on Data Analysis and Visual \nAnalytics (FODAVA) research initiative is in its third year, developing \nstate-of-the-art approaches for analyzing massive and complex data sets.\n In September 2010, Edmond Chow joined the Georgia Tech School of \nComputational Science and Engineering as an associate professor to \ncontinue his work applying numerical and discrete algorithms to \nsimulated physical and scientific systems such as microbiology and \nquantum chemistry as part of Georgia Tech\u0027s new Institute for Data and \nHigh Performance Computing (GTIDH). \u003C\/p\u003E\n\t\u003Cp\u003EGeorgia Tech is making the investments in personnel and \ninfrastructure required to be positioned competitively alongside the \nnation\u0027s top HPC institutions. The Institute will continue to support \nresearch and educational initiatives that push the boundaries of \ntechnological capabilities and broaden the reach of computing \ninnovation.\u003C\/p\u003E\u003Cp\u003EPlease visit Booth 1561 at the SC10 show in New Orleans, LA November 13-19, 2010.\u003C\/p\u003E\n\n\u003Cp\u003E\u0026nbsp;\u003C\/p\u003E","summary":null,"format":"limited_html"}],"field_subtitle":[{"value":"Strategic initiatives in heterogeneous systems, massive parallelism and massive data analytics lead the way"}],"field_summary":[{"value":"\u003Cp\u003EGeorgia Tech displays high performance computing issues such as \nsustainability, reliability and massive data computation November 13-19,\n 2010 at SC10 in New Orleans, LA.\u003C\/p\u003E","format":"limited_html"}],"field_summary_sentence":[{"value":"Strategic initiatives in heterogeneous systems, massive parallelism and massive data analytics lead the way."}],"uid":"27310","created_gmt":"2010-11-10 16:06:07","changed_gmt":"2016-10-08 03:07:42","author":"David Terraso","boilerplate_text":"","field_publication":"","field_article_url":"","dateline":{"date":"2010-11-10T00:00:00-05:00","iso_date":"2010-11-10T00:00:00-05:00","tz":"America\/New_York"},"extras":[],"hg_media":{"46038":{"id":"46038","type":"image","title":"Klaus building","body":null,"created":"1449174347","gmt_created":"2015-12-03 20:25:47","changed":"1475894409","gmt_changed":"2016-10-08 02:40:09","alt":"Klaus building","file":{"fid":"190089","name":"tuv62996.jpg","image_path":"\/sites\/default\/files\/images\/tuv62996_0.jpg","image_full_path":"http:\/\/tlwarc.hg.gatech.edu\/\/sites\/default\/files\/images\/tuv62996_0.jpg","mime":"image\/jpeg","size":40752,"path_740":"http:\/\/tlwarc.hg.gatech.edu\/sites\/default\/files\/styles\/740xx_scale\/public\/images\/tuv62996_0.jpg?itok=UK4jSwpw"}}},"media_ids":["46038"],"groups":[{"id":"1183","name":"Home"}],"categories":[{"id":"153","name":"Computer Science\/Information Technology and Security"}],"keywords":[{"id":"3427","name":"High performance computing"},{"id":"702","name":"hpc"},{"id":"167565","name":"sc10"},{"id":"11229","name":"vetter"}],"core_research_areas":[],"news_room_topics":[],"event_categories":[],"invited_audience":[],"affiliations":[],"classification":[],"areas_of_expertise":[],"news_and_recent_appearances":[],"phone":[],"contact":[{"value":"\u003Cp\u003EStefany Sanders\u003Cbr \/\u003E\nCollege of Computing\u003Cbr \/\u003E\n404-312-6620\u003C\/p\u003E","format":"limited_html"}],"email":["stefany@cc.gatech.edu"],"slides":[],"orientation":[],"userdata":""}}}