Beginning in the 3rd week of class you will have critiques due for papers assigned. Please use the following guidelines for the critique. See COURSE OUTLINE for the number of critiques required each week. Each critique should be done for 1 of the assigned papers (not critiquing multiple papers in a single critique) You are allowed to miss 1 critique without loss of grade.
Critique Guidelines
Structure of the report:
First half: identify one idea in the paper that you feel is a major contribution or a major limitation, explain it and discuss why it is important Second half: describe one idea of yours that builds on the paper and expand on that idea as much as possible (e.g. why is the idea interesting/important?, is it feasible?, etc).
Do not feel constrained to present new ideas solely in the domain of on-chip networks. If you believe this work could be extended to another type of interconnection network, feel free to draw connections between domains.
Please limit the report to 1 page of text. This limit will force you to write more concisely. Utilize the page well.
Focus on depth, not breadth
I am not interested in seeing three major contributions/limitations and ten new ideas, each of which is described in one sentence. I want to see just one idea, explained well.
Your report is not meant to be a summary or an abstract of the paper.
It is not meant to show me that you skimmed the paper.
I want to see how you think and how you can express an opinion or idea in a clear, concise and well-justified manner.
Please provide a printed copy of your report in class at the beginning of lecture.
This is important. I want to see your thoughts before the paper is discussed in class with the whole group.
Origin
This format has been used with success in previous courses. Specifically this form has been used by Prof. Kutulakos at CS. Examples of well done reports (in the field of computer vision) can be found here: link.
Last modified: Sun Jan 2 12:21:09 EST 2011