BEGIN:VCALENDAR
PRODID:-//Microsoft Corporation//Outlook MIMEDIR//EN
VERSION:1.0
BEGIN:VEVENT
DTSTART:20121114T203000Z
DTEND:20121114T211500Z
LOCATION:Ballroom-EFGH
DESCRIPTION;ENCODING=QUOTED-PRINTABLE:ABSTRACT: In 2008, PGI designed the PGI Accelerator programming model and began work on an implementation to target heterogeneous X64 host + NVIDIA GPU systems. In November 2011, Cray, NVIDIA and CAPS Entreprise joined with PGI to refine and standardize directive-based GPU and accelerator programming with the introduction of the OpenACC API. This presentation will discuss three aspects of this language design evolution. We will describe how the programming model changed over time to take advantage of the features of current accelerators, while trying to avoid various performance cliffs. We describe advantages and problems associated with committee-designed languages and specifications.  Finally, we describe several specific challenges related to the implementation of OpenACC for the current generation of targets, and how we solved them in the PGI Accelerator compilers.
SUMMARY:Design, Implementation and Evolution of High Level Accelerator Programming
PRIORITY:3
END:VEVENT
END:VCALENDAR