BEGIN:VCALENDAR
PRODID:-//Microsoft Corporation//Outlook MIMEDIR//EN
VERSION:1.0
BEGIN:VEVENT
DTSTART:20101117T193000Z
DTEND:20101117T200000Z
LOCATION:391-392
DESCRIPTION;ENCODING=QUOTED-PRINTABLE:ABSTRACT: General-Purpose Graphics Processing Units (GPGPUs) are promising parallel=0A  platforms for high performance computing.  The CUDA (Compute Unified Device=0A  Architecture) programming model provides improved programmability for general =0A  computing on GPGPUs.  However, its unique execution model and memory model=0A  still pose significant challenges for developers of efficient GPGPU code.=0A  This paper proposes a new programming interface, called OpenMPC, which builds=0A  on OpenMP to provide an abstraction of the complex CUDA programming model,=0A  and offers high-level controls of the involved parameters and optimizations.=0A  We have developed a fully automatic compilation and user-assisted tuning=0A  system supporting OpenMPC. In addition to a range of compiler transformations=0A  and optimizations, the system includes tuning capabilities for generating,=0A  pruning and navigating the search space of compilation variants. Our results =0A  demonstrate that OpenMPC offers both programmability and tunability. Our  system achieves 88% of the performance of the hand-coded CUDA programs.
SUMMARY:OpenMPC: Extended OpenMP Programming and Tuning for GPUs
PRIORITY:3
END:VEVENT
END:VCALENDAR