Tim’s Suggestions

0. From Ray Chen: how many paper published also mean how efficient you use your time

1. I always think that CAD is the best training for a ECE Ph.D. You will need to learn hardware, software and algorithm. You can basically do anything after you get out of school.

2. I don’t know how to evaluate a Ph.D. in circuit design. If it’s only technical

3. Ph.D. is not judge by how good programming skill you have or how well you can implement things. You need to show intellectual depth on something.

4. I always encourage students to take more classes especially in their first two years of research. Some of them thought that they know everything before taking classes!

5. Paper survey is indeed important. A paper will get rejected directly if an important paper is not cited or compared.

6. You need to know what other people have done and problems are remain unsolved. This is why paper survey is indeed important.

7. The idea of a screening exam is to guarantee the quality of a Ph.D. The methods may be debatable. In an oral exam, they also want to test on you how you will react under pressure and in a short period of time. You will need to defense your concept like a lawyer. This scenario is usually seen in a conference.

8. You are the one who choose to be here and learn. It’s up to you to decide how much you want to learn. No one can force you to make this decision.

9. The quality of a person is evaluated in a constant manner and in every detail.

10. EDA industry is pretty saturated. It is because they have a wrong business model. They do not charge per chip sold but sell it as a software.

11. Experiment result is just to support your arguments and algorithm. The idea is the concept behind the result. Why is your method better than others? What kind of improvement have you done. No one will be interested in seeing just your result. Being able to explain the insights behind your result is more important.

12. Many Ph.D. may have a chance to summarize their work into a new field or categories, but they do not do it. It is really a pity.

13. ITC, DAC, DATE, ICCAD are all top tier conferences. One paper accepted in a conference among this list may be rejected in another. It is because they have different focus and they organize in a different way. A conference that can attract 800+ papers is not a bad conference at all (talking about DATE)

14. If your audience do not understand you, it’s your fault, not your audience’s fault.

15. Top-tier papers goes to conferences. Some journals are only for second tier researches and often not first hand results.

16. In a good conference, there are OK papers and good papers. Usually only 20% of them may be of high quality. Getting accepted is not an easy task.

17. Need to identify what is people’s main concern. Identify what’s of real concern is really important. If your work is interesting to people, it is more appreciated.

18. A proposal is different from a paper. You need to have a “global picture” and “vision” in a proposal. A proposal that lacks these two parts are not good proposal.

19. In the end, people who take initiatives have more chances to success.

20. (Prof. Li from NCU), You do not study for your adviser, you study for yourself in a Ph.D. program

21. (Prof. Li from NCU), Research does not require extremely high talent. Almost everyone has the same level of talent and as long as you work hard, think and critique frequently, a Ph.D. is not a hard stuff.

22. You should know how to formulate the problem you are going to solve. Once a problem is formulated, you also need to express it clear enough so that other people know what is the problem you are going to solve. Both these are not easy tasks and they take efforts.

23. As a leader for the project, I do not have more implementation knowledge than you. The only thing I do is to ask questions. I believe asking good questions usually can lead to good results.

24. Many people have wrong focus on the EDA issue. For example, some researchers are working on something in nanotechnology that are not even proven to be working. For a new technology or new method, what people are interested in is to first show that this idea does work, and then CAD issues come in. Going the other way around is waste of time and moeny.

25. Always implement your idea first before doing further analysis. On the way of implementing your idea, you not only will prove the feasibiliy of your approach but can also identify potential problems in the approach and can formulate analysis on that regard.

26. When asking questions, people do not focus on what will work, they ask what will not work. For example, using carbon-nanotube for interconnect may be very beneficial, but people will tell you ten more other reasons that technique can not be applied.

27. Always have the application in mind so that your method can be more useful.

28. Writing a paper is as a journalist writing a report. The title is the key and more intriguing idea of all. The first paragraph is the summary of the whole report. Although the article may seem to be long, it has only one (or few) key ideas. The following paragraphs of the article is to support the key message conveyed from the title and the first paragraph. Usually, the interview of different people, the evidences from different perspectives are included. These are to support the key idea and to persuade readers better.

29. A good researcher can write a great article for an OK idea. While bad researchers, for many times, a great idea is poorly expressed.

30. Designing proper experiments is hard but a must-do process. Many scientists got Nobel Prize because of their great experiments.

31. When writing a paper, showing that your approach is ‘different’ from others is a must, but demonstrating your approach is ‘better’ than others is even more important.

32. The valuable “products” of an university is its students and the papers.

33. If I saw a paper published in a second tier conference/journal and never follow up in a first tier conference/journal, I would think there are something wrong with their theory so that they can’t achieve good results. I won’t even read these papers.

34. Many activities done at school may not be economical at all, but that is the necessary learning process for training.

35. If I don’t care about your paper, why do I bother spending my time and effort on correcting it. You need to learn from what I have modified and see why I modify in that way.

36. Many students spent a lot of paragraph in the introduction introducing and motivating the background and spent only only sentence on their own appraoch, assuming people will read the rest of the paper. But in reality, people don’t!

37. If I cannot see what’s new in your presentation, I easily get bored and I will walk away immediately. You need to show the meat at the very beginning.

38. Presentation is audience centric so does paper writing. Show others what the important contribution you have at the very beginning.

39. English writing is different from Chinese writing. You need to put conclusion at the very beginning.

40. If you cannot solve the problem, change the problem by changing the assumption, changing the target, etc.

41. People like to work with smart and hard-working people.

42. If you don’t have a schedule, you don’t even know how much have you delayed.

43. Novelty, interesting, and usefulness are the three main criteria to select a good research topic.

44. When you start writing the experimental result, don’t just list all the things that you did. Start with: we executed the experiment to answer the following questions: … then people will grasp the point why you want to do the experiments.

45. Always show the agenda so people know what you are going to do for the meeting.

46. If you respect other’s time, other’s most likely will respect your time.

47. We all tend to judge others. If I think someone is not trustworthy, I will not pay attention to this person or his/her work.

48. You evaluate a person’s personality when there’s a conflict and a person’s quality when there’s pressure.

More to come…

Responses

  1. ^^
    interesting thoughts

  2. […] Tim’s Suggestions […]

  3. ITC and DATE are not top tier conferences. Top tier conferences don’t have on average 1-2 citations per paper. In addition, neither their PCs or people publishing there indicate very high quality

    DAC and ICCAD on the other hand have > 5 citations per paper and are definitely top tier.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: