BEGIN:VCALENDAR PRODID:-//Microsoft Corporation//Outlook MIMEDIR//EN VERSION:1.0 BEGIN:VEVENT DTSTART:20111114T163000Z DTEND:20111115T010000Z LOCATION: DESCRIPTION;ENCODING=QUOTED-PRINTABLE:ABSTRACT: GPUs allow advanced high performance computing with lower budgets,=0Aaffecting scientific computing from very high-end supercomputing systems=0Adown to the departmental and personal level. Learn how to make=0Aeffective use of the available performance on GPUs. This tutorial uses=0Aprogressively more complex examples to motivate performance=0Aoptimization, and will include several hands-on exercises. Attendees=0Awill receive instructions on how to pre-load exercises and GPU=0Adevelopment tools on their Windows, Linux, or MacOS laptops prior to the=0Atutorial. Many of the exercises can be done even on laptops without a=0AGPU. The morning session will focus on optimizing CUDA C, OpenCL and=0ACUDA Fortran programs. The afternoon session will continue with=0Aperformance optimization using directive-based GPU programming,=0Aincluding the PGI Accelerator Programming model and the proposed OpenMP=0AAccelerator directives. SUMMARY:M05: Advanced GPU Computing PRIORITY:3 END:VEVENT END:VCALENDAR