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:20190719T085743Z
LOCATION:HG F 3
DTSTART;TZID=Europe/Stockholm:20190612T163000
DTEND;TZID=Europe/Stockholm:20190612T170000
UID:submissions.pasc-conference.org_PASC19_sess129_msa169@linklings.com
SUMMARY:Automatic Generation of Production-Grade Hybrid MPI-OpenMP Paralle
 l Wave Propagators using Devito
DESCRIPTION:Minisymposium\nComputer Science and Applied Mathematics, Chemi
 stry and Materials, Climate and Weather, Physics, Life Sciences, Engineeri
 ng\n\nAutomatic Generation of Production-Grade Hybrid MPI-OpenMP Parallel 
 Wave Propagators using Devito\n\nLuporini\n\nDevito is a Python-based doma
 in-specific language for implementing <br />high-performance finite differ
 ence partial differential equation solvers. The <br />motivating applicati
 on is exploration seismology where methods such as <br />Full-Waveform Inv
 ersion (FWI) and Reverse-Time Migration (RTM) are used to invert <br />ter
 abytes of seismic data to create images of the earth's subsurface. Even us
 ing modern <br />supercomputers, it can take weeks to process a single sei
 smic survey and create <br />a useful subsurface image. The computational 
 cost is dominated by the numerical <br />solution of wave equations and th
 eir corresponding adjoints. Additionally, <br />the actual set of par
 tial differential equations being solved and their <br />numerical discret
 ization is under constant innovation as increasingly realistic <br />repre
 sentations of the physics are developed, further ratcheting up the cost of
  <br />practical solvers. By embedding a domain-specific language &nd
 ash; based on the symbolic <br />mathematics library SymPy – wi
 thin Python and exploiting sophisticated compiler <br />technology, D
 evito makes it possible to quickly develop highly-optimized fini
 te <br />difference solvers. In this talk, we will present the l
 anguage, compiler architecture, <br />applications and performance of
  Devito, with emphasis on automatically generated <br />MPI-Open
 MP production-grade wave propagators for high-frequency RTM. Fun
 damental <br />features such as boundary conditions and interpolation
  will also be discussed.
END:VEVENT
END:VCALENDAR