BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:GPU Architecture Challenges for Throughput Computing
DTSTART:20110208T151500
DTSTAMP:20260501T095122Z
UID:ab6c05e16e0e3497e2d2fb4df7789c609ca962b4fc896716fe6aad9e
CATEGORIES:Conferences - Seminars
DESCRIPTION:Dr Tor M. Aamodt\nThere is growing interest in using graphics 
 processors (GPUs) to accelerate non-graphics\, throughput-oriented softwar
 e.  GPUs have the potential to significantly improve performance per unit 
 cost versus general purpose CPUs for applications containing significant p
 arallelism. This talk will provide a brief introduction to GPU Computing a
 nd summarize some of the key challenges contemporary GPUs present to softw
 are developers beyond the need to write parallel code.  Then it will descr
 ibe recent research on microarchitecture mechanisms for improving branch h
 andling\, scalable memory access scheduling\, and on-chip interconnection 
 networks in accelerator architectures. This talk will also briefly describ
 e GPGPU-Sim\, a simulator capable of running GPU Computing applications wr
 itten in CUDA and OpenCL developed to enable architecture research on thes
 e challenges.
LOCATION:BC 02 https://plan.epfl.ch/?room==BC%2002
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR