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:20260522T150111Z
LOCATION:C2/3/4 Ballroom
DTSTART;TZID=America/Chicago:20181115T083000
DTEND;TZID=America/Chicago:20181115T170000
UID:submissions.supercomputing.org_SC18_sess324_post259@linklings.com
SUMMARY:Which Architecture Is Better Suited for Matrix-Free Finite-Element
  Algorithms: Intel Skylake or Nvidia Volta?
DESCRIPTION:Martin Kronbichler (Technical University Munich), Momme Allale
 n and Martin Ohlerich (Leibniz Supercomputing Centre), and Wolfgang A. Wal
 l (Technical University Munich)\n\nThis work presents a performance compar
 ison of highly tuned matrix-free finite element kernels from the finite el
 ement library on different contemporary computer architectures, NVIDIA V10
 0 and P100 GPUs, an Intel Knights Landing Xeon Phi, and two multi-core Int
 el CPUs (Broadwell and Skylake).  The algorithms are based on fast integra
 tion on hexahedra using sum factorization techniques.  For small problem s
 izes, when all data fits into CPU caches, Skylake is very competitive with
  Volta. For larger sizes, however, the GPU holds an advantage of approxima
 tely a factor of three over Skylake, because all architectures operate in 
 the memory-bandwidth limited regime. A detailed performance analysis contr
 asts the throughput-oriented character of GPUs versus the more latency-opt
 imized CPUs for the scenario of high-order finite element computations.\n\
 nRegistration Category: Tech Program Reg Pass, Exhibits Reg Pass\n\n
END:VEVENT
END:VCALENDAR