BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Europe/Stockholm
X-LIC-LOCATION:Europe/Stockholm
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20190719T085744Z
LOCATION:HG EO Nord
DTSTART;TZID=Europe/Stockholm:20190613T195000
DTEND;TZID=Europe/Stockholm:20190613T215000
UID:submissions.pasc-conference.org_PASC19_sess179_post114@linklings.com
SUMMARY:CSM03 - A Parallel Space-Time Boundary Element Method
DESCRIPTION:Poster\n\n\nCSM03 - A Parallel Space-Time Boundary Element Met
 hod\n\nMerta, Zapletal, Dohr\n\nParallel space-time algorithms are conside
 red to be one of the possibilities how to improve scalability on curr
 ent multi-petascale and future exascale supercomputers, as opposed to the 
 time-stepping algorithms where the parallelization is usually limited to t
 he spatial dimension. On the other hand, the high computational intensity 
 of the boundary element method (BEM) makes it well-suited for state-of-the
 -art multi- and many-core processors equipped with wide SIMD registers. Pa
 rallel space-time BEM therefore promises to maximize the utilization 
 of resources of the large-scale heterogeneous clusters. Here we present th
 e space-time boundary element method for the heat equation which employs a
 ll possible levels of parallelization, including SIMD vectorization and sh
 ared memory parallelization using OpenMP, and distributed memory paralleli
 zation based on the scheme utilizing cyclic graph decompositions for load 
 balancing. To evaluate the performance of the proposed method, we pre
 sent scalability experiments on multiple CPU architectures.
END:VEVENT
END:VCALENDAR