This seminar is held by the
Electrical/Computer Engineering Department
at University of Delaware. Professor Stéphane Zuckerman is overseeing the Fall 2015 class.
Semiconductor technologies and physical constraints have already forced
microprocessor vendors to go from single-core to multi-core technology.
The same fundamental limitations of the underlying hardware now forces
them to go beyond multi-core and enter the many-core era, which implies
a renewal of the parallel computing field.
This 800 graduate research seminar course will be focused on the
challenges and opportunities in parallel computing systems. The course
will be focused on the design and implementation of future high-end
computer systems -- from multiple peta-scale up to exa-scale computers.
The main challenges are centered on the definition and specification of
parallel program execution models that are sound, scalable and
programmable for the intended applications.
The course will cover a few well selected advanced subject areas as well
as their practical impact on algorithms, architectures and software
design. These may include (but not restricted to) the following:
concurrency models, dataflow & multithreading, execution models,
performance evaluation of execution models, and dynamic runtime &
The class will be organized into reading and discussion groups and the
studies will be guided by key lectures, reading assignment, group
presentations and discussion, and in-depth exploitation of topics for
Time and Location
Tuesdays 5:45 PM - 08:15 PM
Location: Gore Hall 218
First meeting: September 1st
Prof. Stéphane Zuckerman
201B Evans Hall
121 Evans Hall