For the last several weeks of the semester you should be working on a
final project. At the end of the semester you will both give a 20
minute talk on your project, and turn in a final report. However,
along the way there are various intermediate deadlines.
This page describes those deadlines and what's expected for each of
them. I will be adding on to this page as the various deadlines
approach. At any point in the semester you should feel free to make an
appointment to talk to me about the project (should you find yourself
looking for direction, advice, feedback, motivation, etc).
Your final project can be either a survey paper (ie, literature
review), or something more hands-on. In coming up with a topic
you might start by thinking about whether there's
an application that you're already interested in solving, if
there's a topic that you've always wanted to know more about (maybe
inspired by the papers we've read in this class?),
if you enjoyed writing MPI code, and so on.
The following table gives the reading assignments for 11/13. Everyone
is responsible for about 35 pages of reading (the number in parentheses
after the title of each paper gives the page count), and at least one article
is one that you selected on your own topic (so hopefully you've already
read that one).
|Paper || link || Tim || Alex || Kristen || Carl || Frank || Danny || Charles |
| A Grid-Enabled MPI: Message passing in
heterogeneous distributed computing systems (9) ||
|| X || X || || X || X || X ||
| Razor: A low-power pipeline based on circuit-level timing speculation (12)||
|| || || X || || X || ||
| Power-efficient value speculation for high-performance microprocessors (8)||
|| || || || X || X || ||
| Temporal difference learning and TD-Gammon (11) ||
|| X || X || || || || ||
| Co-evolution in the successful learning of backgammon strategy (16)||
|| || || X || || || || X
| n-Body simulations using message passing parallel computers (5)||
|| || || || || || X ||
| A practical comparison of n-body algorithms (16)||
|| X || X || || || || ||
| Understanding the efficiency of GPU algorithms for matrix-matrix multiplication (5)||
|| || || X || || X || X || X
| Reducing the total bandwidth of a sparse unsymmetric matrix (17)||
|| || || || X || || X || X
In addition to including all the other things you've been noting in
your paper summaries all semester, write a few sentences on whether you
think it's a good paper and why.
Put each summary on a separate sheet of paper and bring 2 copies of
to class on 11/13 (one copy is for me and one is for the person who's
doing the project on that topic).
You should each be prepared to speak for about 10 minutes in class
on 11/15 about the current status of your project (things you've
discovered/accomplished/etc) and about what you hope to have done
by the end of the semester. Feel free to ask for feedback/comments/suggestions
from the class.
You should also be prepared to ask questions about other people's
projects, particularly since each of you will have read related
papers for at least two other projects.
The order for 11/15 is:
The presentations will take place in class on 11/29 and 12/4.
The order for the presentations is:
The final report is due by 5PM on 12/6, although I won't complain as long
as I have it by midnight on Friday, 12/8.