DATE: Thursday, October 18, 2001
TIME: Noon - 1 pm
PLACE: Newell Simon Hall 3305
SPEAKER:
Yale
Patt
University of Texas at Austin
TITLE:
Moore's Law: Is it Time to
Just Give Up and Work on Easier Problems?
ABSTRACT:
Moore's Law has been misused to predict just about every metric imagineable
when it comes to future microprocessors. Regarding technology, there seems
to be no problem -- very soon one billion transistors on a die, and geometries
that allow a dozen gate delays within a 10GHz clock. Regarding performance,
on the other hand, there are plenty of nay-sayers who are recommending
we pack it in and give up. In this talk, we will explore what the naysayers
are saying and whether or not we should listen. We will also discuss some
of the avenues that could prove fruitful in keeping us on the curve of
Moore's Law with respect to performance. Per usual, the talk will undoubtedly
be laced with digressions, based on questions from the audience or some
random thoughts that occur to the speaker.
BIO:
Yale Patt is a teacher at The University of Texas at Austin. He has obtained
the right set of appropriate degrees from reputable universities, and
more than enough awards for his research and teaching. He is best known
for having had Garth Gibson, Greg Ganger and Dave Nagle in class. Fortunately
for him, no two of them were in his class at the same time. More detail
is available on his web page: http://www.ece.utexas.edu/~patt/.
SDI / LCS Seminar Questions?
Karen Lindenfelser, 86716, or visit www.pdl.cmu.edu/SDI/