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:20181221T160903Z
LOCATION:C2/3/4 Ballroom
DTSTART;TZID=America/Chicago:20181115T083000
DTEND;TZID=America/Chicago:20181115T170000
UID:submissions.supercomputing.org_SC18_sess324_post114@linklings.com
SUMMARY:The NAStJA Framework: Non-Collective Scalable Global Communication
 s
DESCRIPTION:Poster\nTech Program Reg Pass, Exhibits Reg Pass\n\nThe NAStJA
  Framework: Non-Collective Scalable Global Communications\n\nBerghoff, Kon
 dov\n\nIn recent years, simulations in various areas of science and engine
 ering have proven to be very useful.  To efficiently deploy simulation cod
 es on current and future high-performance computer systems, high node leve
 l performance, scalable communication and the exclusion of unnecessary cal
 culations are an absolute must when developing new solvers.\n\nWe have int
 roduced the NAStJA framework, a block-based MPI parallel solver for algori
 thms, based on regular grid methods, i.e., stencil codes.  NAStJA has a dy
 namic block adaptation, which modifies the calculation domain around the r
 egion in which the calculation is currently taking place.  The creation an
 d deletion of blocks are autonomously managed within local neighborhoods. 
  Collective all-gather communication is avoided by using a multi-hop netwo
 rk to distribute information across the entire domain that greatly improve
 s the application scaling.  In this contribution, we present applications 
 that can benefit from this adaptive method and scaling tests demonstrating
  the excellent scalability.
URL:https://sc18.supercomputing.org/presentation/?id=post114&sess=sess324
END:VEVENT
END:VCALENDAR