Do not mistake- almost all IITs have direct admit for M.Tech. CSE. Only IITK and IITD have interviews. So, my post is meant for research/project interviews. There are plenty of good interview experiences given by previous year people. Still I wanted to say something new.
Last year and even before I have seen many people clearing interviews and usually the same set clears them in most IITs. This might be a surprise because every place more than 200 people attend and only like 10-15 gets selected. But that is the truth- other people are simply not good enough. So, my post is to try and help them.
Many IITs do have a filter test where GATE level questions are asked with more importance to questions in TIFR syllabus (theory). I guess you all know how to clear this.
Now, for the main interviews professors look for those who are interested in research. They do not need someone who is trying hard to answer they mugged up. Somethings they like are follows:
- University/college topper - shows a student works hard
- Some national level achievement like math olympiad, NTSC etc. - shows he/she is skillful
- Publications - very important and if its an international one highly valued
- Good project experience -- research oriented projects have higher value
- Knowledge of academically relevant tools - in many labs they have simulators or similar tools. Good working experience on them is a plus.
At B.Tech. level it is not easy to even read a good publication. And we cannot change your academic history also. But rest of the things we can change. So, i suggest to see last year ranks (once you know your marks) and target a particular IIT and a particular lab there. Then prepare to reach there, even if you attempt at other places. i.e., your focus must be on a particular lab/area. I uploaded a slide on performance monitoring which is relevant for compiler/architecture/HPC areas. Like this there are many tools which are relevant for each domain. I cannot help you on other areas because I simply have no idea. Only way is to google- navigate via IIT profs/labs pages. Once you decide your area and have a lab in mind things are very narrowed and you can proceed easily. This is the main step and most difficult too.
This is not something many people do. But research is meant to be new. If you go to interview thinking you are going to be selected just because others are worse than you, it is never going to happen. Only way to be selected is with quality - absolute selection and not relative.