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:20260522T150115Z
LOCATION:D165
DTSTART;TZID=America/Chicago:20181112T115000
DTEND;TZID=America/Chicago:20181112T121000
UID:submissions.supercomputing.org_SC18_sess161_ws_pmbss101@linklings.com
SUMMARY:miniVite: A Graph Analytics Benchmarking Tool for Massively Parall
 el Systems
DESCRIPTION:Sayan Ghosh (Washington State University), Mahantesh Halappana
 var and Antonino Tumeo (Pacific Northwest National Laboratory), and Ananth
  Kalyanaraman and Assefaw Gebremedhin (Washington State University)\n\nBen
 chmarking of high performance computing systems can help provide critical 
 insights for efficient design of computing systems and software applicatio
 ns. Although a large number of tools for benchmarking exist, there is a la
 ck of representative benchmarks for the class of irregular computations as
  exemplified by graph analytics. We therefore propose miniVite as a repres
 entative graph analytics benchmark tool to test a variety of distributed-m
 emory systems. Graph clustering, popularly known as community detection, i
 s a prototypical graph operation used in numerous scientific computing and
  analytics applications. The goal of clustering is to partition a graph in
 to clusters (or communities) such that each cluster consists of vertices t
 hat are densely connected within the cluster and sparsely connected to the
  rest of the graph. Modularity optimization is a popular technique for ide
 ntifying clusters in a graph. Efficient parallelization of modularity opti
 mization-based algorithms is challenging. One successful approach was conc
 eived in Vite, a distributed-memory implementation of the Louvain method t
 hat incorporates several heuristics. We develop miniVite as a representati
 ve but simplified variant of Vite, to serve as a prototypical graph analyt
 ics benchmarking tool. Similar to many graph algorithms, miniVite is chara
 cterized by irregular communication patterns, high communication to comput
 ation ratios, and load imbalances among participating processes, thus maki
 ng it a representative benchmarking tool. \n\nUnlike some graph-based meth
 ods such as breadth-first search and betweenness centrality, miniVite repr
 esents highly complex computational patterns stressing a variety of system
  features. This can in turn help provide crucial insight for codesign of f
 uture computing systems. We believe that miniVite will serve as a platform
  for benchmarking systems and design communication primitives that will be
  applicable to a broad set of irregular computing applications as well as 
 a platform for the design of efficient graph algorithms.\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)\n\n
END:VEVENT
END:VCALENDAR