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_post135@linklings.com
SUMMARY:PHY01 - A Fully Traceless Cartesian Multipole Formulation for the 
 Distributed Fast Multipole Method
DESCRIPTION:Poster\n\n\nPHY01 - A Fully Traceless Cartesian Multipole Form
 ulation for the Distributed Fast Multipole Method\n\nColes, Bieri\n\nThe f
 ast multipole method (FMM) is used to efficiently approximate long-range i
 nteraction potentials found in diverse fields from molecular dynamics to a
 strophysics. This algorithm is particularly important now in the era of ex
 ascale computing because of its O(n) scaling and low communication overhea
 d. In many applications, the potential is Taylor-expanded using Cartesian 
 tensors, which are manipulated with a collection of operator functions to 
 compute the total potential and its derivatives. It is well-known that if 
 the tensors are written in a traceless form, the total computation and mem
 ory requirements are significantly reduced. However, writing the correspon
 ding equations for the FMM operators in an easy to understand, and easy to
  implement, form is challenging. We present a clear formulation of the equ
 ations that uses fully traceless tensors and demonstrate a software genera
 tor that automatically produces and symbolically optimises the operator co
 de. We demonstrate the efficiency of these routines, which have been imple
 mented in the Polaris(MD) molecular dynamics package, with a biophysical a
 pplication. We observe a 20% speed-up and a 47% memory reduction over a no
 n-traceless implementation.
END:VEVENT
END:VCALENDAR