Computer Engineering Seminar

Computer Architecture Research: Some Criticisms and Some Opportunities

Yale Patt

At a time when there is unprecedented opportunity for the computer
architecture community, we are instead hearing negative comments from
some of our "leaders" such as:

* we are not a serious discipline,
* everything that is worth doing has been done, and
* we are in serious need of revitalization.

I will first respond to each of these. Then I will point out our unique
opportunities and responsibilities — both the obvious ones and the not-so-
obvious ones. The obvious ones involve the microprocessor of the year 2016,

and what I think we will have to do to be successful in that venture. The not-so-
obvious ones are in areas that we have traditionally left to others.

Yale Patt is a teacher at UT Austin, where he holds the Ernest Cockrell Jr.
Centennial Chair in Engineering. He enjoys teaching 400 students at a time
in his Intro to Computing, and 40 students at a time in his advanced
course in Microarchitecture. In fact, the 400 student freshman course
originated at Michigan as EECS 100 more than ten years ago where he
co-developed it with Professor Kevin Compton. With respect to his research,
he and his students have been responsible for HPS, the Two-level Branch
Predictor, Helper Threads, and Wish Branches, among other things.
He currently supervises 11 graduate students and has consulted for
the microprocessor industry since it began. He has earned appropriate
from reputable universities and has received more than enough awards for
his research and teaching. For those who want more information, you can

Sponsored by