BEGIN:VCALENDAR
PRODID:-//Microsoft Corporation//Outlook MIMEDIR//EN
VERSION:1.0
BEGIN:VEVENT
DTSTART:20121114T001500Z
DTEND:20121114T020000Z
LOCATION:East Entrance
DESCRIPTION;ENCODING=QUOTED-PRINTABLE:ABSTRACT: We use a particle-based method to simulate incompressible=0Aflows, where the Fast Multipole Method (FMM) is used=0Ato accelerate the calculation of particle interactions. The=0Amost time-consuming kernelsthe Biot-Savart equation and=0Astretching term of the vorticity equationare mathematically=0Areformulated so that only two Laplace scalar potentials are=0Aused instead of six, while automatically ensuring divergencefree=0Afar-field computation. Based on this formulation, and=0Aon our previous work for a scalar heterogeneous FMM=0Aalgorithm, we develop a new FMM-based vortex method=0Acapable of simulating general flows including turbulence on=0Aheterogeneous architectures. Our work for this poster focuses=0Aon the computation perspective and our implementation can=0Aperform one time step of the velocity+stretching for one billion=0Aparticles on 32 nodes in 55.9 seconds, which yields 49.12=0ATflop/s
SUMMARY:Scalable Fast Multipole Methods for Vortex Element Methods
PRIORITY:3
END:VEVENT
END:VCALENDAR
BEGIN:VCALENDAR
PRODID:-//Microsoft Corporation//Outlook MIMEDIR//EN
VERSION:1.0
BEGIN:VEVENT
DTSTART:20121114T001500Z
DTEND:20121114T020000Z
LOCATION:East Entrance
DESCRIPTION;ENCODING=QUOTED-PRINTABLE:ABSTRACT: We use a particle-based method to simulate incompressible=0Aflows, where the Fast Multipole Method (FMM) is used=0Ato accelerate the calculation of particle interactions. The=0Amost time-consuming kernelsthe Biot-Savart equation and=0Astretching term of the vorticity equationare mathematically=0Areformulated so that only two Laplace scalar potentials are=0Aused instead of six, while automatically ensuring divergencefree=0Afar-field computation. Based on this formulation, and=0Aon our previous work for a scalar heterogeneous FMM=0Aalgorithm, we develop a new FMM-based vortex method=0Acapable of simulating general flows including turbulence on=0Aheterogeneous architectures. Our work for this poster focuses=0Aon the computation perspective and our implementation can=0Aperform one time step of the velocity+stretching for one billion=0Aparticles on 32 nodes in 55.9 seconds, which yields 49.12=0ATflop/s
SUMMARY:Scalable Fast Multipole Methods for Vortex Element Methods
PRIORITY:3
END:VEVENT
END:VCALENDAR