CS2750 – Topics in Parallel and Distributed Computing
Spring 2014
Tuesday and Thursday 1:00-2:20
CIT 368
Schedule
Starting Points
Course Information
Grades and Assignments
Course Information
Simply put, the goal of this course is to hone certain core skills essential for success as a researcher or practitioner of Computer Science. Imagine that some day your boss, or thesis advisor, or principal investor, or president, or spouse says “you have a week to please bring me up to date on state-of-the-art Foozle research”. Failure will be a CLM (“career limiting move”). What will you do?
The goal of this course is to take a snapshot of current research topics in distributed and concurrent computing. We will start with papers published in the principal conferences in this area.
Progress at the forefront of research is often incremental: one researcher publishes a paper posing a question or claiming a result, and a sequence of follow-on papers improve the result or alter the question. For this reason, we will organize our approach around the idea of clusters of papers. A cluster consists of one primary paper, the one to read if you can read only one, together with two or three secondary papers. The primary paper may have been the first to formulate the problem or technique, or it may have provided the best solution to the problem, or perhaps it is simply the most readable.
Schedule
Slides
Phoenix++
Grades and
Assignments
Participation (10%
of course grade)
Research papers are often poorly written, sometimes make exaggerated or misleading claims, and occasionally contain errors or major ambiguities (imagine that!). I expect students to contribute to the discussion by asking questions, making observations, and subjecting material to critical scrutiny. These skills will be useful in any area of science.
Most important:
the course won’t be any fun without
lively participation from the studio audience.
Presentations
(40% of course grade)
Students will work in teams of two, and each team should plan to make about four presentations. In consultation with the instructor, each team will:
- Identify a topic,
- Identify an primary paper (see above), and
- Identify two or three secondary papers.
The team will give a presentation on the topic, with an emphasis on:
- The basic problem or technique,
- A critical evaluation of the primary paper,
- The context and depth provided by secondary papers, and
- Open research questions.
Teams are advised, but not required, to show their presentations to the instructor before the presentation.
Each presentation will have 80 minutes (one class period). At least one week before the presentation, the team will post the primary paper to the web page. At the time of the presentation, the team will deliver to the instructor some version of the presentation suitable for posting on the course web page.
Depending on the course enrollment, teams may have to present more than once.
Paper Evaluations
(10% of course grade)
A paper evaluation form consists of:
- Your name
- The paper name
- Summarize the paper (no more than five sentences)
- Most important strengths (no more than three, one sentence each)
- Most important weaknesses (no more than three, one sentence each)
- State one problem or issue left open (no more than three sentences).
Paper evaluations will be graded on a scale of one to three. The default grade is two. Insightful reviews get three, and disappointing reviews get one.
Students will email evaluations of primary papers to the instructor (cs275.brown@gmail.com) before the start of the class in which the paper is presented. Late or incomplete evaluations get no credit. Students are required to evaluate at least two-thirds of the primary papers presented.
Presentation
Evaluations (10% of course grade)
You are also required to evaluate presentations. Why? First, if you have to write a review of someone else’s talk, you had better pay attention. Second, if you know that your own talk is being evaluated by the studio audience (not just the instructor), then you may try harder to appeal to them. In the Real World™, when you graduate, you will have to capture the attention of intelligent, well-educated audiences that know little or nothing about your field. Sharpen your skills now.
A presentation evaluation form must contain the following fields:
Name: Your name
Presenters: who’s talking?
Vision: how well did the presenters explain why the area matters?
Style: did the presenters mumble, fail to make eye contact, speak too quickly, too slowly, or what?
Exposition: were the PowerPoint slides too busy, too ugly, or just right?
Q&A: How well did the presenters seem to know the material? Were they honest about admitting when they don’t know something?
Comments: anything else you would like to say.
Presentation evaluations will be graded on a scale of one to three. The default grade is two. Insightful evaluations get three, and disappointing evaluations get one. Evaluations for talks where I suspect the reviewer was not physically present get zero.
Presentation evaluations are intended to be helpful. It is OK to be frank (otherwise what’s the point?) but be polite (no matter how you are provoked). I will merge and edit presentation evaluations and forward them to the presenters. Your evaluations will be kept anonymous, and I retain the right to edit or suppress intemperate or inappropriate comments.
Students will email evaluations of presentations the
instructor (cs275.brown@cs.brown.edu)
before Friday
Project (30% of
course grade)
The final project requirements are the same as for the presentation, except that
- You work alone, not in a team, and
- You write a term paper, not a presentation.
- You may choose to work on the same topic as your presentation, but the papers covered should not be the same
Consult the instructor if there is any question.
Missive:
Introduction
(slides)
Some Starting
Points:
· PODC 2013
· DISC 2013
· Transact 2013
· ASPLOS 2013
· EuroSys
2013
· OSDI
2012
· PPoPP 2013
· SPAA 2013
Some of these pages have links to the papers, and some have titles only. You may need to use search engines or contact the authors.
Instructor
