BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Dynamic Scheduling at Large Scale
DTSTART:20120725T150000
DTSTAMP:20260406T042135Z
UID:b4ac5993492e6492e5e8f4ef33de410a8dc0793ddd7ac4187bdebc64
CATEGORIES:Conferences - Seminars
DESCRIPTION:Christos Kozyrakis\nStanford University\nMulti-core chips will
  soon include hundreds of cores\, support thousands of hardware threads\, 
 and feature deep memory hierarchies with non-uniform latency characteristi
 cs. In such systems\, the latency and energy overheads of remote memory ac
 cesses will dwarf the latency and energy overheads of computation. To maxi
 mize efficiency\, we must schedule parallel computations in a manner that 
 optimizes for locality\, while maintaining load balance and minimizing sch
 eduling overheads.\n\nThis talk will present the work on dynamic schedulin
 g for large-scale multi-core systems at the Pervasive Parallelism Lab (PPL
 ) in Stanford University. First\, the broad potential of locality-aware sc
 heduling and the importance of using high-level information from the devel
 oper or the programming model in achieving efficient execution will be dis
 cussed. Second\, GRAMPS\, a scheduling and runtime system for pipeline-par
 allel programs that optimizes memory behavior while performing fine-grain 
 dynamic load balancing with low overhead\, will be presented. Even on toda
 y's multi-core chips\, GRAMPS outperforms the commonly used scheduling app
 roaches such as task-stealing\, GPGPU\, and static streaming schedulers. T
 hird\, simple hardware support that allows for the development of low-over
 head\, software-mostly runtime systems for fine-grain parallelism that sca
 le efficiently to hundreds of hardware threads will be presented. Finally\
 , directions for future work in dynamic resource management in large-scale
  parallel systems will be discussed. 
LOCATION:BC 410 https://plan.epfl.ch/?room==BC%20410
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR