Times report on Graduate School applications

As expected the number of people applying to graduate school has gone up. What is surprising is that the number of people taking the GRE has gone down... Article here.

Questions from Harsh

Hi...first i would like to thank you for regularly posting on blogs and helping people like us who want to get into US.

I am currently in my final year of computer science and engineering. My scores are not way too good, my aggregate till now is 63%. I want to do a MS from a good university. By good , I mean top 50 universities. So, what else should i do to compensate for
my relatively lesser grades?. Actually, I am interested in Phd and research but, it is too time consuming and money is the major factor here to think about. So, at first I am going to complete my MS.

Also, considering the future, which is the most feasible field to choose for MS. I am particularly interested in Data mining, computer architecture and design, operating systems and machine learning. Other interests include artificial intelligence and nanotechnology. So which should i choose, or should i go for a generic degree in ECE?. I dream of working at google/Nvidia.

I have been a constant reader of IEEE papers and some works of professors from Berkeley and MIT. I also am thinking of some research ideas , but I need to get into one of the top 50 universities so as to make the best out of my career. So do let me
know your thoughts about this.

thanks
Harsh

Thanks for visiting my blog Harsh. I'm glad that I could help you in any way. You do sound pretty well-informed about most things. I think you should be able to make it into a top 50 university. 63% is a respectable score for at least the University of Pune.

Google
For a job at Google, data mining & operating systems probably are good areas of specialization. Google is working on Google Android actively, and pushing it into both the netbook and cellphone markets. So that is definitely a growth area for their company. I personally feel voice recognition (or natural language processing) is also a good area to be in. Microsoft made some announcements yesterday about how they imagine the computing world will be like 10 years from now. They talked about how we will all have personal assistants. If you've heard of a service called GOOG-411, it allows users to conduct searches providing voice input. Image Processing is probably a good expertise for them to deal with all those images for Google Maps.

NVIDIA
For a job at NVIDIA, computer architecture, VLSI design, operating systems, compilers are good areas of research. A big growth area for NVIDIA is CUDA. It is an API/Architecture for parallel computing, which exposes a C extension allowing users to execute parallel tasks on the GPU. On some apps users have reported a speed-up by a factor of 100. Even Intel is trying to get users to program for multi-core, which is a big area for research.

To improve your odds
- Like I have said in the past paper presentation contests, extra projects are a good way to get an edge over others. Academic scores are important but you can compensate at least a little by displaying some practical usage of your skills.
- For a computer science major, if you could come up with a simple idea for something like a useful website, a CUDA based application speeding up something simple like JPEG encoding or video encoding, contributing to some open source project like Linux are good ways to get an edge over your competition.
- If you are reading at least some IEEE journals you already have a head start.

Other pointers
- These are just my opinions.
- You'll have to focus on a couple of aims for now or at least focus on one aim for your applications. When you are at the university you can try out a few different fields and then choose what interests you. For applications I think it is best to express interest in one area.
- Just a pointer, if you do plan to go for a PhD you'll have to do an MS thesis. Keep in mind that if you transfer from one university to another for your PhD, all your MS credits may not get counted.
- Nanotechnology is a little bit of a hard-sell for CS majors because it is at the boundary of Electrical Engineering and Physics for most parts. It may not be the best use of your CS skills.

Words starting with the letter F

Quizzes by Quibblo.com