FRS 001N — Sec. 001
— (1 unit) — CRN 55880 — W 3:10 – 4:00pm — 203
Wellman
Parallel and Distributed Computing
Instructor: Roland Faller, Department of Chemical Engineering &
Materials Science, College of Engineering
Description: Many problems nowadays can only be
addressed reliably
with heavy computing resources, such resources may be in the form
of large computing
facilities (e.g. National Leadership Class Computing Facility) or in the form
of very distributed resources (e.g. Folding at Home, SETI at home).
In the latter
case regular personal computers are used in their “free time” by
running a special screen saver to solve pressing scientific problems like the
folding patterns of proteins which are essential to find new drugs
for various
diseases or the existence of extraterrestrial intelligence.
Everybody can participate
by donating computer time; the only prerequisite is an internet connection.
Without such an arrangement these problems may not be solved in the near term
due to financial constraints. On the other hand large national
computing facilities
have been very successful in solving problems like the humane genome project.
This class will familiarize students with the concepts of
distributed computing;
explain its capabilities and limitations. The students will get to understand
the basic concepts of scaling up a problem, algorithmic principles
and infrastructural
needs. A focus will be important examples to which the students can directly
connect and about which they may have heard in the media. There
will be no specific
computer programming performed or discussed but the students will be educated
in general terms. The instructor has a small computing cluster for his group;
this cluster will be used as an example for a small project.
Format: Seminar
will meet one session (one hour) per week, there is no text book.
Students will
on their own research the existence of large facilities and
distributed projects
(mainly by using the internet) and compare them. All sessions are
interactive;
students are expected to prepare themselves for class by reading
material prior
to the class. Students will become aware of the technology behind
large computational
resources. The often perceived initial hesitance to approach such
systems should
be reduced and at the same time unrealistic expectations be damned. Students
are expected to write a term paper about one facility/project of
their choice,
the level of detail would be that the interested layman can
understand it. Grading:
2/3 on Term Paper 1/3 on general participation.
About the Instructor: Professor Faller is a member
of the faculty
in the Department of Chemical Engineering & Materials Science.
His research
interests include the computer modeling of soft materials,
especially polymers
and biomembranes. Current projects include the interplay of cell
membranes with
alcohols and sugars as well as the fundamental understanding of glasses. He
currently teaches a course on thermodynamics.