BEGIN:VCALENDAR
PRODID:-//Microsoft Corporation//Outlook MIMEDIR//EN
VERSION:1.0
BEGIN:VEVENT
DTSTART:20121115T230000Z
DTEND:20121115T233000Z
LOCATION:255-BC
DESCRIPTION;ENCODING=QUOTED-PRINTABLE:ABSTRACT: In this paper, we describe the design and performance of GRAPE-8=0Aaccelerator processor for gravitational N-body simulations.  It is=0Adesigned to evaluate gravitational interaction with cutoff between=0Aparticles. The cutoff function is useful for schemes like TreePM or=0AParticle-Particle Particle-Tree, in which gravitational force is=0Adivided to short-range and long-range components.  A single GRAPE-8=0Aprocessor chip integrates 48 pipeline processors. The effective number=0Aof floating-point operations per interaction is around 40. Thus the=0Apeak performance of a single GRAPE-8 processor chip is 480 Gflops. A=0AGRAPE-8 processor card houses two GRAPE-8 chips and one FPGA chip for=0APCI-Express interface. The total power consumption of the board is=0A46W. Thus, theoretical peak performance per wattage is 20.5 Gflops/W.=0AThe effective performance of the total system, including the host=0Acomputer, is around 5Gflops/W. This is more than a factor of two=0Ahigher than the highest number in the current Green500 list.
SUMMARY:GRAPE-8 - An Accelerator for Gravitational N-Body Simulation with 20.5GFLOPS/W Performance
PRIORITY:3
END:VEVENT
END:VCALENDAR
BEGIN:VCALENDAR
PRODID:-//Microsoft Corporation//Outlook MIMEDIR//EN
VERSION:1.0
BEGIN:VEVENT
DTSTART:20121115T230000Z
DTEND:20121115T233000Z
LOCATION:255-BC
DESCRIPTION;ENCODING=QUOTED-PRINTABLE:ABSTRACT: In this paper, we describe the design and performance of GRAPE-8=0Aaccelerator processor for gravitational N-body simulations.  It is=0Adesigned to evaluate gravitational interaction with cutoff between=0Aparticles. The cutoff function is useful for schemes like TreePM or=0AParticle-Particle Particle-Tree, in which gravitational force is=0Adivided to short-range and long-range components.  A single GRAPE-8=0Aprocessor chip integrates 48 pipeline processors. The effective number=0Aof floating-point operations per interaction is around 40. Thus the=0Apeak performance of a single GRAPE-8 processor chip is 480 Gflops. A=0AGRAPE-8 processor card houses two GRAPE-8 chips and one FPGA chip for=0APCI-Express interface. The total power consumption of the board is=0A46W. Thus, theoretical peak performance per wattage is 20.5 Gflops/W.=0AThe effective performance of the total system, including the host=0Acomputer, is around 5Gflops/W. This is more than a factor of two=0Ahigher than the highest number in the current Green500 list.
SUMMARY:GRAPE-8 - An Accelerator for Gravitational N-Body Simulation with 20.5GFLOPS/W Performance
PRIORITY:3
END:VEVENT
END:VCALENDAR