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_post109@linklings.com
SUMMARY:CSM14 - Sparse Approximate Matrix Multiplication in a Fully Recurs
 ive Distributed Task-Based Parallel Framework
DESCRIPTION:Poster\n\n\nCSM14 - Sparse Approximate Matrix Multiplication i
 n a Fully Recursive Distributed Task-Based Parallel Framework\n\nArtemov\n
 \nWe consider a parallel implementation of approximate multiplication of l
 arge matrices with decay. Such matrices arise in computations related to e
 lectronic structure calculations and some other fields of science. The ori
 ginal algorithm for the sparse approximate multiplication was suggested by
  [N. Bock and M. Challacombe, An optimized sparse approximate matrix multi
 ply for matrices with decay]. An implementation done using the Chunks and 
 Tasks programming model and library [E.H. Rubensson and E. Rudberg, Chunks
  and Tasks: A programming model for parallelization of dynamic algorithms]
  is presented and discussed. We describe a two-level approach, where the o
 uter one operates with tasks in parallel, while the inner one performs act
 ual computations within tasks. The implementation of the algorithm is appl
 ied to large chemical systems with more than 10^6 atoms. We found out that
  it is competitive to another popular approach, which performs truncation 
 of small blocks before multiplication. A comparison between these two meth
 ods in terms of performance on a model problem is done. The method is then
  applied to real matrices arising in quantum chemistry.
END:VEVENT
END:VCALENDAR