BEGIN:VCALENDAR
PRODID:-//Microsoft Corporation//Outlook MIMEDIR//EN
VERSION:1.0
BEGIN:VEVENT
DTSTART:20121114T223000Z
DTEND:20121115T000000Z
LOCATION:155-F
DESCRIPTION;ENCODING=QUOTED-PRINTABLE:ABSTRACT: Many-core devices, such as GPUs, are widely adopted as part of high-performance, distributed computing. In every cluster setting, efficient resource management is essential to maximize the cluster utilization and the delivered performance, while minimizing the failure rate. Currently software components for many-core devices, such as the CUDA, provide limited support to concurrency and expose low-level interfaces which do not scale well and are therefore not suitable to cluster and cloud environments. In our research, we aim to develop runtime technologies that allow managing tasks in large-scale heterogeneous clusters, so to maximize the cluster utilization and minimize the failures exposed to end users. As manufacturers are marketing a variety of many-core devices with different hardware characteristics and software APIs, we will propose a unified management component hiding the peculiarities of the underlying many-core devices to the end users. Our study will focus on algorithms and mechanisms for scheduling and fault recovery.
SUMMARY:An Efficient Runtime Technology for Many-Core Device Virtualization in Clusters
PRIORITY:3
END:VEVENT
END:VCALENDAR