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:20181221T160727Z
LOCATION:D161
DTSTART;TZID=America/Chicago:20181112T113000
DTEND;TZID=America/Chicago:20181112T115000
UID:submissions.supercomputing.org_SC18_sess158_ws_lasalss120@linklings.co
 m
SUMMARY:Communication Reduced Multi-Timestep Algorithm for Real-Time Wind 
 Simulation on GPU-Based Supercomputers
DESCRIPTION:Workshop\nAlgorithms, Heterogeneous Systems, Resiliency, Works
 hop Reg Pass\n\nCommunication Reduced Multi-Timestep Algorithm for Real-Ti
 me Wind Simulation on GPU-Based Supercomputers\n\nOnodera, Idomura, Ali, S
 himokawabe\n\nWe develop a communication reduced multi-time-step (CRMT) al
 gorithm for a Lattice Boltzmann method (LBM) based on a block-structured a
 daptive mesh refinement (AMR). This algorithm is based on the temporal blo
 cking method, and can improve computational efficiency by replacing a comm
 unication bottleneck with additional computation. The proposed method is i
 mplemented on an extreme scale airflow simulation code CityLBM, and its im
 pact on the scalability is tested on GPU based supercomputers, TSUBAME and
  Reedbush. Thanks to the CRMT algorithm, the communication cost is reduced
  by ∼ 64%, and weak and strong scalings are improved up to ∼ 2
 00 GPUs. The obtained performance indicates that real time airflow simulat
 ions for about 2km square area with the wind speed of ∼ 5m/s is feas
 ible using 1m resolution.
URL:https://sc18.supercomputing.org/presentation/?id=ws_lasalss120&sess=se
 ss158
END:VEVENT
END:VCALENDAR