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:20211207T055412Z
LOCATION:Online
DTSTART;TZID=America/Chicago:20211114T154000
DTEND;TZID=America/Chicago:20211114T154500
UID:submissions.supercomputing.org_SC21_sess328_ws_whpc103@linklings.com
SUMMARY:Early Career Lighting Talks - A GPU Parallel Algorithm for Computi
 ng Morse-Smale Complexes
DESCRIPTION:Workshop\n\nEarly Career Lighting Talks - A GPU Parallel Algor
 ithm for Computing Morse-Smale Complexes\n\nSubhash\n\nThe Morse-Smale com
 plex is a well studied topological structure that represents the gradient 
 flow behavior between critical points of a scalar function. It supports mu
 lti-scale topological analysis and visualization of feature-rich scientifi
 c data. Several parallel algorithms have been proposed towards the fast co
 mputation of the 3D Morse-Smale complex. Its computation continues to pose
  significant algorithmic challenges. In particular, the non-trivial struct
 ure of the connections between the saddle critical points are not amenable
  to parallel computation. This paper describes a fine grained parallel alg
 orithm for computing the Morse-Smale complex and a GPU implementation gMSC
 . The algorithm first determines the saddle-saddle reachability via a tran
 sformation into a sequence of vector operations, and next computes the pat
 hs between saddles by transforming it into a sequence of matrix operations
 . Computational experiments show that the method achieves up to 8.6x speed
 up over current shared memory implementations. Our individual algorithms f
 or marking saddle reachability and path counting achieve speedups of up to
  577.7x and 5.4x respectively. The paper also presents a comprehensive exp
 erimental analysis of different steps of the algorithmic pipeline and repo
 rts on their contribution towards runtime performance. Finally, it introdu
 ces a CPU based data parallel algorithm for simplifying the Morse-Smale co
 mplex via iterative critical point pair cancellation which achieves a spee
 dup of up to 13.2x.\n\nTag: Online Only, Career Development, Diversity Equ
 ity Inclusion (DEI), Education and Training and Outreach, HPC Community Co
 llaboration\n\nRegistration Category: Workshop Reg Pass
END:VEVENT
END:VCALENDAR