BEGIN:VCALENDAR
PRODID:-//Microsoft Corporation//Outlook MIMEDIR//EN
VERSION:1.0
BEGIN:VEVENT
DTSTART:20101117T181500Z
DTEND:20101117T191500Z
LOCATION:390
DESCRIPTION;ENCODING=QUOTED-PRINTABLE:ABSTRACT: The Cray XMT is an unconventional high-performance computer platform well suited for data analytics and knowledge discovery.  The system’s large shared-memory, latency tolerant processors, and fine-grained hardware synchronization enable highly irregular memory access patterns and algorithms to scale.  Researchers at national laboratories and universities are using the machine to develop large- scale applications in bioinformatics, network analysis, community detection, text processing, natural language understanding, and blog analysis.  Other multithreading researchers are developing software tools including libraries, compilers, performance monitors, and hardware simulators.=0A=0AWith the emergence of many-core processors capable of maintaining multiple software threads per core and terabyte shared memory nodes, conventional systems may soon be able to emulate many of the hardware features found now only in the XMT.  Consequently, debate exists in the community whether future multi-threaded architectures should be primarily hardware based, software based, or some hybrid.  A hardware implementation will be costly, but is guaranteed to deliver exceptional performance for a class of applications not served by traditional HPC vendors.  Conversely, a software implementation will enjoy the cost benefits and legacy infrastructure of commodity hardware, but will performance be acceptable?=0A=0AThis Birds-of-a-Feature will bring together the massive multithreading community to discuss results, current status, and future machine designs driven by application requirements.  The meeting will feature talks on application performance, demonstrations of programming tools, and position papers on expectations of next-generation multithreaded systems.
SUMMARY:The XMT User Community Birds-of-a-Feather
PRIORITY:3
END:VEVENT
END:VCALENDAR