Here is a general template for paper summaries for the readings we will do in the course. 

Write around 1-2 pages (10/11 point font) summarizing the paper. Please do not reproduce the abstract, or copy any parts of the paper (text or figs) verbatim. The purpose is for me to understand how far you have understood the concepts in the paper. It will help with calibrating the future papers and the level at which this class should be taught.

You can cover some/all of the questions here in your summaries. Remember that they are guidelines, and you should use your judgement to figure out which of these questions pertain to the paper you are reading. Feel free to cover more points than provided here if you like.

 - What problem is the paper solving?

- What is the motivation for solving that problem?

- What was the state-of-the-art in solving that problem when the paper was written?

- What is the contribution of this paper over the state of the art (at the time it was written)?

- Describe the solution proposed by the paper. You don't have to provide the psuedocode or flowchart, but can refer to it (eg. In Fig 5, a block diagram is shown with 3 phases...) while describing the algo. You can either use a precise description of the algo (hard) or provide an example of its working (easier and more concrete).

-Provide a working example for the algorithm. You can either use an example from the paper, or your own example to illustrate the algorithm/technique proposed by the paper. If you are using your own example, please indicate that.  

- What is the complexity of the solution presented? You don't have to provide a formal complexity analysis, but you should provide an intuition of where the solution faces scalability issues (eg. this is exponential in X where X is ..., hence in the worst case it will blow up when X is large, but practically still be feasible)

- How do the authors present an evaluation of their solution? (Type of designs, size of designs, comparisons against benchmarks, metrics)

- Is the evaluation satisfactory in terms of benchmarks used, the metrics recorded and does it justify the claims made in contributions?

- Is the demonstrated evaluation sufficient to determine scaling to modern systems? Does it show practical feasibility?

- Do a Google search and check whether this technology is used in modern designs and EDA tools. If so, in what form? Name the tools that implement this technology today.

- What did you find most interesting (strength) in this paper? If you did not find anything interesting, give reasons why not (weaknesses).

- If you had to solve the problem solved by the paper, how would you approach it? (It is fine to answer with a commonplace or non-research technique. Idea is to get thinking of research problems.

- Can you think of any other research questions/problems that the solution in the paper could be applied to? Again, you don't have to conduct research on this topic. Idea is to get you thinking about open ended/creative ways to apply solutions in research.