CSE788.I11: Reliability and Variability Aware Architectures
- Winter 2009
Instructor: Radu
Teodorescu
Office:
783 Dreese Labs
Email:
{teodores} in CSE at Ohio
State
Class days, time and location: TR, 0930-1048, DL 0317
Call # 21607-4
Outline
Reliability of
computing devices is one of the "grand challenges" the
computing industry will be facing over the next decade and
beyond. The growth in microprocessor performance, which has
driven our industry for many years, is now threatened by
the reliability challenges brought on by the very
aggressive technology scaling.
This course explores the state of the art in architectural
and system innovations that enable the continued
improvement in performance in the face of these challenges.
We will discuss new architecture and system designs that
help detect, mitigate and tolerate a variety of error
classes including transient errors, faults due to process
variation, wearout and aging, design errors, etc.
We will also talk about ways in which architecture can help
make software more reliable by providing support for
efficient software debugging.
The goal of this class is to learn the state of the art,
but also to come up with new ways of designing reliable
systems from unreliable components.
One page flier
Reading list, assignments, projects and class managements
in Carmen.