BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Chicago
X-LIC-LOCATION:America/Chicago
BEGIN:DAYLIGHT
TZOFFSETFROM:-0600
TZOFFSETTO:-0500
TZNAME:CDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0500
TZOFFSETTO:-0600
TZNAME:CST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260522T150127Z
LOCATION:D165
DTSTART;TZID=America/Chicago:20181112T085900
DTEND;TZID=America/Chicago:20181112T173000
UID:submissions.supercomputing.org_SC18_sess161@linklings.com
SUMMARY:The 9th International Workshop on Performance Modeling, Benchmarki
 ng, and Simulation of High-Performance Computer Systems (PMBS18)
DESCRIPTION:The PMBS18 workshop is concerned with the comparison of high-p
 erformance computer systems through performance modeling, benchmarking or 
 through the use of tools such as simulators.\n\nThe aim of this workshop i
 s to bring together researchers, from industry and academia, concerned wit
 h the qualitative and quantitative evaluation and modeling of high-perform
 ance computing systems. Authors are invited to submit novel research in al
 l areas of performance modeling, benchmarking and simulation, and we welco
 me research that brings together current theory and practice. We recognize
  that the coverage of the term performance has broadened to include power 
 consumption and reliability, and that performance modeling is practiced th
 rough analytical methods and approaches based on software tools and simula
 tors.\n\nEvaluating SLURM Simulator with Real-Machine SLURM and Vice Versa
 \n\nHaving a precise and a fast job scheduler model that resembles the rea
 l-machine job scheduling software behavior is extremely important in the f
 ield of job scheduling. The idea behind SLURM simulator is preserving the 
 original code of the core SLURM functions while allowing for all the advan
 tages of...\n\n\nAna Jokanovic, Marco D'Amico, and Julita Corbalan (Barcel
 ona Supercomputing Center)\n---------------------\nEvaluating the Impact o
 f Spiking Neural Network Traffic on Extreme-Scale Hybrid Systems\n\nAs we 
 approach the limits of Moore's law, there is increasing interest in non-Vo
 n Neuman architectures such as neuromorphic computing to take advantage of
  improved compute and low power capabilities. Spiking neural network (SNN)
  applications have so far shown very promising results running on a numb..
 .\n\n\nNoah Wolfe and Mark Plagge (Rensselaer Polytechnic Institute), Misb
 ah Mubarak (Argonne National Laboratory), Christopher D. Carothers (Rensse
 laer Polytechnic Institute), and Robert B. Ross (Argonne National Laborato
 ry)\n---------------------\nApproximating a Multi-Grid Solver\n\nMulti-gri
 d methods are numerical algorithms used in parallel and distributed proces
 sing. The main idea of multi-grid solvers is to speed up the convergence o
 f an iterative method by reducing the problem to a coarser grid a number o
 f times. Multi-grid methods are widely exploited in many application ...\n
 \n\nValentin Le Fèvre (ENS Lyon) and Leonardo Bautista-Gomez, Osman Unsal,
  and Marc Casas (Barcelona Supercomputing Center)\n---------------------\n
 Benchmarking Machine Learning Methods for Performance Modeling of Scientif
 ic Applications\n\nPerformance modeling is an important and active area of
  research in high-performance computing (HPC). It helps in better job sche
 duling and also improves overall performance of coupled applications. Suff
 iciently rich analytical models are challenging to develop, however, becau
 se of interactions betw...\n\n\nPreeti Malakar, Prasanna Balaprakash, Venk
 atram Vishwanat, Vitali Morozov, and Kalyan Kumaran (Argonne National Labo
 ratory)\n---------------------\nImproving MPI Reduction Performance for Ma
 nycore Architectures with OpenMP and Data Compression\n\nMPI reductions ar
 e widely used in many scientific applications and often become the scaling
  performance bottleneck. When performing reductions on vectors, different 
 algorithms have been developed to balance messaging overhead and bandwidth
 . However, most implementations have ignored the effect of si...\n\n\nHong
 zhang Shan and Samuel Williams (Lawrence Berkeley National Laboratory) and
  Calvin Johnson (San Diego State University)\n---------------------\nExplo
 ring and Quantifying How Communication Behaviors in Proxies Relate to Real
  Applications\n\nProxy applications, or proxies, are simple applications m
 eant to exercise systems in a way that mimics real applications (their par
 ents). However, characterizing the relationship between the behavior of pa
 rent and proxy applications is not an easy task. In prior work, we present
 ed a data-driven meth...\n\n\nOmar Aaziz and Jeanine Cook (Sandia National
  Laboratories), Jonathan Cook (New Mexico State University), and Courtenay
  Vaughan (Sandia National Laboratories)\n---------------------\nA Metric f
 or Evaluating Supercomputer Performance in the Era of Extreme Heterogeneit
 y\n\nWhen acquiring a supercomputer, it is desirable to specify its perfor
 mance using a single number. For many procurements, this is usually stated
  as a performance increase over a current generation platform, for example
  machine A provides 10 times greater performance than machine B. The deter
 mination ...\n\n\nBrian Austin, Chris Daley, Douglas Doerfler, Jack Deslip
 pe, Brandon Cook, Brian Friesen, Thorsten Kurth, Charlene Yang, and Nichol
 as Wright (Lawrence Berkeley National Laboratory)\n---------------------\n
 Workshop Lunch (on your own)\n---------------------\nDeep Learning at Scal
 e on Nvidia V100 Accelerators\n\nThe recent explosion in the popularity of
  Deep Learning (DL) is due to a combination of improved algorithms, access
  to large datasets and increased computational power. This had led to a pl
 ethora of open-source DL frameworks, each with varying characteristics and
  capabilities. End users are then lef...\n\n\nRengan Xu, Frank Han, and Qu
 y Ta (Dell EMC)\n---------------------\nWorkshop Morning Break\n----------
 -----------\nAlgorithm Selection of MPI Collectives Using Machine Learning
  Techniques\n\nAutotuning is a well established method to improve software
  performance for a given system, and it is especially important in High Pe
 rformance Computing. The goal of autotuning is to find the best possible a
 lgorithm and its best parameter settings for a given instance. Autotuning 
 can also be applied...\n\n\nSascha Hunold (Technical University Wien) and 
 Alexandra Carpen-Amarie (Fraunhofer Institute for Industrial Mathematics)\
 n---------------------\nIntroduction - The 9th International Workshop on P
 erformance Modeling, Benchmarking, and Simulation of High-Performance Comp
 uter Systems (PMBS18)\n\nThe PMBS18 workshop is concerned with the compari
 son of high-performance computing systems through performance modeling, be
 nchmarking or through the use of tools such as simulators. We are particul
 arly interested in research which reports the ability to measure and make 
 tradeoffs in software/hardwar...\n\n\nSteven A. Wright (University of York
 ), Stephen A. Jarvis (University of Warwick), and Simon D. Hammond (Sandia
  National Laboratories)\n---------------------\nUnified Cross-Platform Pro
 filing of Parallel C++ Applications\n\nTo address the great variety of ava
 ilable parallel hardware architectures (CPUs, GPUs, etc.), high-performanc
 e applications increasingly demand cross-platform portability.  While unif
 ied programming models like OpenCL or SYCL provide the ultimate portabilit
 y of code, the profiling of applications in...\n\n\nVladyslav Kucher, Flor
 ian Fey, and Sergei Gorlatch (University of Münster)\n--------------------
 -\nminiVite: A Graph Analytics Benchmarking Tool for Massively Parallel Sy
 stems\n\nBenchmarking of high performance computing systems can help provi
 de critical insights for efficient design of computing systems and softwar
 e applications. Although a large number of tools for benchmarking exist, t
 here is a lack of representative benchmarks for the class of irregular com
 putations as ...\n\n\nSayan Ghosh (Washington State University), Mahantesh
  Halappanavar and Antonino Tumeo (Pacific Northwest National Laboratory), 
 and Ananth Kalyanaraman and Assefaw Gebremedhin (Washington State Universi
 ty)\n---------------------\nIs Data Placement Optimization Still Relevant 
 on Newer GPUs?\n\nModern supercomputers often use Graphic Processing Units
  (or GPUs) to meet the evergrowing demands for energy efficient high perfo
 rmance computing. GPUs have a complex memory architecture with various typ
 es of memories and caches, such as global memory, shared memory, constant 
 memory, and texture me...\n\n\nMd Abdullah Shahneous Bari (Stony Brook Uni
 versity); Larisa Stoltzfus (University of Edinburgh); Pei-Hung Lin, Chunhu
 a Liao, and Murali Emani (Lawrence Livermore National Laboratory); and Bar
 bara Chapman (Stony Brook University, Brookhaven National Laboratory)\n---
 ------------------\nWorkshop Afternoon Break\n---------------------\nAutom
 ated Instruction Stream Throughput Prediction for Intel and AMD Microarchi
 tectures\n\nAn accurate prediction of scheduling and execution of instruct
 ion streams is a necessary prerequisite for predicting the in-core perform
 ance behavior of throughput-bound loop kernels on out-of-order processor a
 rchitectures. Such predictions are an indispensable component of analytica
 l performance mo...\n\n\nJan Laukemann (University of Erlangen-Nuremberg);
  Julian Hammer (University of Erlangen-Nuremberg, Erlangen Regional Comput
 ing Center); Johannes Hofmann (University of Erlangen-Nuremberg); Georg Ha
 ger (University of Erlangen-Nuremberg, Erlangen Regional Computing Center)
 ; and Gerhard Wellein (University of Erlangen-Nuremberg)\n\nTag: Benchmark
 s, Parallel Programming Languages, Libraries, and Models, Performance, Sim
 ulation\n\nRegistration Category: Workshop Reg Pass\n\nSession Chair: Stev
 en A. Wright (University of York, England)
END:VEVENT
END:VCALENDAR