{"159711":{"#nid":"159711","#data":{"type":"event","title":"Ph.D. Defense of Dissertation: Paul Bryan","body":[{"value":"\u003Cp\u003E\u003Cstrong\u003EPh.D. Defense of Dissertation\u003Cbr \/\u003E\u003C\/strong\u003E\u003C\/p\u003E\u003Cp\u003ETitle: \u003Cstrong\u003EAccelerating Microarchitectural Simulation via Statistical Sampling Principles\u003C\/strong\u003E\u003Cbr \/\u003E\u003Cbr \/\u003EPaul D. Bryan\u003Cbr \/\u003ECollege of Computing\u003Cbr \/\u003EGeorgia Institute of Technology\u003Cbr \/\u003E\u0026nbsp;\u003Cbr \/\u003E\u003Cbr \/\u003EDate: Friday, October 12, 2012\u003Cbr \/\u003ETime: 8am \u2013 11am\u003Cbr \/\u003ELocation: Klaus 3402\u003Cbr \/\u003E\u0026nbsp;\u003Cbr \/\u003E\u003Cbr \/\u003E\u003Cstrong\u003ECommittee:\u003C\/strong\u003E\u003C\/p\u003E\u003Cul\u003E\u003Cli\u003EProf. Thomas M. Conte (Advisor, Colleges of Computing \u0026amp; Engineering)\u003C\/li\u003E\u003Cli\u003EProf. Milos Prvulovic (College of Computing)\u003C\/li\u003E\u003Cli\u003EProf. George Riley (College of Engineering)\u003C\/li\u003E\u003Cli\u003EProf. Sudhakar Yalamanchili (College of Engineering)\u003C\/li\u003E\u003Cli\u003EDr. Gabe Loh (Principal Researcher, AMD Research)\u003C\/li\u003E\u003C\/ul\u003E\u003Cp\u003E\u0026nbsp;\u003Cbr \/\u003E\u003Cstrong\u003EAbstract:\u003C\/strong\u003E\u003Cbr \/\u003EThe design and evaluation of computer systems relies heavily upon simulation.\u0026nbsp; Simulation is also a major bottleneck in the iterative design process.\u0026nbsp; Applications that may be executed natively on physical systems in a matter of minutes may take weeks or months to simulate.\u0026nbsp; As designs incorporate increasingly higher numbers of processor cores, it is expected that the times required to simulate future systems will become an even greater issue.\u0026nbsp;\u0026nbsp; Simulation exhibits a tradeoff between speed and accuracy. By basing experimental procedures upon known statistical methods, the simulation of systems may be dramatically accelerated while retaining reliable methods to estimate error.\u003Cbr \/\u003E\u0026nbsp;\u003Cbr \/\u003EThis thesis focuses on the acceleration of simulation through statistical processes.\u0026nbsp; The first two techniques discussed in this thesis focus on accelerating single-threaded simulation via cluster sampling.\u0026nbsp; Cluster sampling extracts multiple groups of contiguous population elements to form a sample.\u0026nbsp; This thesis introduces techniques to reduce sampling and non-sampling bias components, which must be reduced for sample measurements to be reliable.\u0026nbsp; Non-sampling bias is reduced through the Reverse State Reconstruction algorithm, which removes ineffectual instructions from the skipped instruction stream between simulated clusters.\u0026nbsp; Sampling bias is reduced via the Single Pass Sampling Regimen Design Process, which guides the user towards selected representative sampling regimens. Unfortunately, the extension of cluster sampling to include multi-threaded architectures is non-trivial and raises many interesting challenges.\u0026nbsp; Overcoming these challenges will be discussed.\u0026nbsp; This thesis also introduces thread skew, a useful metric that quantitatively measures the non-sampling bias associated with divergent thread progressions at the beginning of a sampling unit.\u0026nbsp; Finally, the Barrier Interval Simulation method is discussed as a technique to dramatically decrease the simulation times of certain classes of multi-threaded programs.\u0026nbsp; It segments a program into discrete intervals, separated by barriers, which are leveraged to avoid many of the challenges that prevent multi-threaded sampling.\u003C\/p\u003E","summary":null,"format":"limited_html"}],"field_subtitle":"","field_summary":"","field_summary_sentence":[{"value":"Accelerating Microarchitectural Simulation via Statistical Sampling Principles"}],"uid":"1","created_gmt":"2012-10-04 15:26:33","changed_gmt":"2016-10-08 02:00:28","author":"Jupiter","boilerplate_text":"","field_publication":"","field_article_url":"","field_event_time":{"event_time_start":"2012-10-12T09:00:00-04:00","event_time_end":"2012-10-12T12:00:00-04:00","event_time_end_last":"2012-10-12T12:00:00-04:00","gmt_time_start":"2012-10-12 13:00:00","gmt_time_end":"2012-10-12 16:00:00","gmt_time_end_last":"2012-10-12 16:00:00","rrule":null,"timezone":"America\/New_York"},"extras":[],"groups":[{"id":"47223","name":"College of Computing"},{"id":"50875","name":"School of Computer Science"}],"categories":[],"keywords":[],"core_research_areas":[],"news_room_topics":[],"event_categories":[],"invited_audience":[],"affiliations":[],"classification":[],"areas_of_expertise":[],"news_and_recent_appearances":[],"phone":[],"contact":[{"value":"\u003Cp\u003E\u003Ca href=\u0022mailto:paul.bryan@gatech.edu\u0022\u003EPaul D. Bryan\u003C\/a\u003E\u003C\/p\u003E","format":"limited_html"}],"email":[],"slides":[],"orientation":[],"userdata":""}}}