Abstract
The 2nd International Workshop on Performance Modeling, Benchmarking and Simulation of High Performance Computing Systems (PMBS'11) was held as part of the 24th ACM/IEEE International Conference for High Performance Computing, Networking, Storage and Analysis (SC'11), which took place at the Washington State Convention Center in Seattle on 12-18 November 2011. The SC conference series is the premier international forum for high-performance computing, networking, storage and analysis. The conference is unique in that it hosts a wide range of international participants from academia, national laboratories and industry; this year's conference attracted a record number of attendees, including over 300 exhibitors in the industry's largest HPC technology fair. SC offers a vibrant technical program, which includes technical papers, tutorials in advanced areas, Birds of a Feather sessions (BoFs), panel debates, a doctoral showcase and a number of technical workshops in specialist areas (of which PMBS is one). The focus of the PMBS 11 workshop was comparing high-performance computing systems through performance modeling, benchmarking or the use of tools such as simulators. We were particularly interested in receiving research papers which reported the ability to measure and make tradeoffs in hardware/software co-design to improve sustained application performance. We were also keen to capture the assessment of future systems, for example through work that ensured continued application scalability through peta- and exa-scale systems. The aim of the PMBS'11 workshop was to bring together researchers from industry, national labs and academia, who were concerned with the qualitative and quantitative evaluation and modeling of high-performance computing systems. Authors were invited to submit novel research in all areas of performance modeling, benchmarking and simulation, and we welcomed research that combined novel theory and practice. We also expressed an interest in submissions that included analysis of power consumption and reliability, and were receptive to performance modeling research that madeuse of analytical methods as well as those based on tracing tools and simulators. Technical submissions were encouraged in areas including: performance modeling and analysis of applications and high-performance computing systems; novel techniques and tools for performance evaluation and prediction; advanced simulation techniques and tools; micro-benchmarking, application benchmarking and tracing; performance-driven code optimization; scalability analysis; verification and validation of performance models; benchmarking and performance analysis of novel hardware; performance concerns in software/hardware co-design; tuning and auto-tuning of HPC applications and algorithms; benchmark suites; performance visualization, and real-world case studies. We received an excellent number of submissions for this year's workshop. This meant that we were able to be very selective in those papers that were chosen. The resulting 13 papers show worldwide programs of research committed to understanding application and architecture performance to enable peta-scale computational science. Contributors to the workshop included Lawrence Berkeley National Laboratory, University of Oxford, Imperial College London, EPFL Switzerland, the US National Center for Supercomputing Applications, University of Tennessee, Oak Ridge National Laboratory, Swiss National Supercomputing Centre, RWTH Aachen University, Cray Inc, North Carolina State University, Virginia Tech, Lawrence Livermore National Laboratory and the University of Auckland, amongst others. The PMBS 11 workshop was extremely well attended and we thank the participants for the lively discussion and positive feedback received throughout the workshop. We hope to be able to repeat this success in future years.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call