Research Challenges in GPU Computing

Wednesday, November 26, 2008 - 10:33am

John Owens
Associate Professor, Electrical and Computer Engineering

Date: 12/2/2008

Time: 2:00PM

Location: CS Conference Rm, HFH 1132

Abstract: The recent explosion in programmability and arithmetic
capability in the graphics processor (GPU) has motivated its use for
computationally demanding applications well beyond the traditional
graphics pipeline. In my talk, I will present a high-level view of
the modern GPU’s hardware and programming model, and then discuss
what I see as some of the most important research challenges that are
facing the GPU computing community. When presenting similar material
at other top venues, I’ve often found that the audience has numerous
questions during and after the talk, which I hope will be the case at
UCSB as well. Please bring your toughest and most interesting
questions and I’ll be prepared to answer them as best I can.

Biography: John Owens is an associate professor of electrical and
computer engineering at the University of California, Davis. His
research interests are in GPU computing (GPGPU) and more broadly
commodity parallel hardware and programming models. John earned his
Ph.D. in electrical engineering in 2003 from Stanford University and
his B.S. in electrical engineering and computer sciences in 1995 from
the University of California, Berkeley.

Departmental Host: Fred Chong, Director and Professor Computer
Engineering Program