Stanford EE M.S. vs UMich CSE PhD (and U of T MASc)

[battousai]

Hi all,

I am currently finishing up my undergrad in Canada. I applied to a number of schools for Computer Engineering in the fall (specializing in Computer Architecture). I've gotten some acceptances from schools I applied to, which I'm very happy about, but now of course I have to make a choice.

I applied to M.S. in some places and PhD in others, partly to try my luck and partly because I don't really want to spend another 5 years in school unless I really like what I'm doing. I'm open to doing a PhD, but only if I feel I'd get a real benefit by doing so.

I applied to the EE M.S. at Stanford and got in (no funding), and also got in to the PhD at UMichigan in CSE (they said with financial aid, don't know how much yet). I was also accepted to the University of Toronto's Masters program (which has a mixture of courses and research).

At the end of my schooling, I'd like to go into industry rather than academia. I've got a fair bit of work experience through internships in my undergrad, so I know what it's like. I'd like to pick the degree that opens the most doors for me. I applied to grad school because I wanted to learn more about computer architecture and associated fields to get a better understanding. I also feel that the creative thinking involved in research would be beneficial when designing/implementing systems in industry.

Stanford of course has the brand name that turns heads, and is close to successful Silicon Valley companies. However, I understand that the M.S. is course based, and Masters students don't really get to do research with profs and learn that way (in the words of PhD comics - "come back when you've finished quals"). Living expenses are also said to be quite high, making for an expensive degree that's really just a bunch of extra courses for the most part.

Michigan, while not as big a brand name as Stanford (but still big in computer architecture), has admission to the PhD program which would definitely allow me to do research. The profs there seem to be doing interesting stuff as well (which is why I applied). There's also at least some funding, though I am willing to pay for a good degree if it'll help me in the long run.

Last but not least, the U of T option allows me to stay in Canada and do research - although I'm not sure what the worldwide reputation of U of T is and how it looks to potential employers in Silicon Valley and elsewhere.

The gist is this: I feel that a research-based degree might be more beneficial to me for learning purposes (i.e. teach me more), but the brand name of Stanford might be too good to pass up. Does a Stanford EE Masters have a lot of value at companies in Comp. Arch. (Intel, NVIDIA, etc), or is its value discounted because Stanford admits a lot of Masters students to make money? Does going to Stanford allow one to develop connections that help lead to good jobs after the degree, or is that only on the PhD side?

If I go the Michigan route but get fed up after completing the Masters requirements (which is also just a bunch of courses), I believe I could leave with just the Masters (and that people have done this), but then I feel I'd be burning my bridges so to speak, and toasting any chance of good references from professors in the future. So that's another factor that weighs on my mind with the Michigan decision.

Any advice would be appreciated. If any of my understanding isn't right, please do let me know. Thanks!

[ase_lima]

I can give you advice as a current MASc student at UofT (albeit with a focus in Electronics and not comp arch. specifically) who is likely headed to Stanford for PhD studies.

First of all, UofT's comp. arch. group has lots of top drawer faculty - you may want to look up these people: Paul Chow, Andreas Moshovos, Greg Steffan, Natalie Enright Jerger. Take a look at their research groups and you'll see that they are very productive and have had numerous publications in leading Conferences and Journals (which for an outsider is one of the best indicators of research productivity). As I have lots of friends in the comp. arch. group, I can tell you that they have very close relations with AMD, Intel, IBM, etc. Moreover, if you are willing to step slightly away from computer architecture specifically and look at, say, FPGAs, I can tell you that UofT has perhaps the world's best FPGA research group (they develop FPGA architectures, CAD tools for FPGAs, soft processors for FPGAs, etc.), and is EXTREMELY well connected with the two major FPGA companies (Altera and Xilinx - in fact Altera has a key research and development centre adjacent to the campus). UofT's industry connections, while perhaps not as good as Stanford's, is still very good, and you really don't have to worry about getting a job after you graduate as most of my peers have graduated with very nice jobs.

On the other hand, I'm not too much of a fan of Stanford's Master's program; yes it offers an excellent selection of courses, but the fact that you will likely get no research exposure (which ends up SERIOUSLY hurting your chances to get into a good PhD program afterwards if you are so inclined) and the fact that you have to pay roughly $100k for the 2 years there (whereas at UofT they would be paying you) doesn't make the program to appealing at all.

I think between Stanford Masters and UofT MASc, I would choose UofT MASc mainly because of the possibility of getting to do some high quality research and not being down $100k. The only reason I'm heading off to Stanford for my PhD is because I have already done two degrees at UofT and need a change of scenery.

[CanuckBoy]

I think Stanford is the best destination for you, based on what I've seen and heard. Coming out of undergrad I've head several classmates head down to Stanford for masters (not Comp Eng) and funding didn't seem to be a problem (maybe they're rich?). You should talk to current Stanford students about funding opportunities.

My brother is currently working in California, and worked for various big name companies. He once told me he was surrounded by people from big name schools, one of the names he mentioned was Stanford.

If you're interested in PhD, I don't think having a course-based master from Stanford will hurt your chances as I've had a couple professors who did their master at Stanford and continued on at Stanford for PhD or great schools elsewhere.

It's true a school will have connections to local industries, but few schools have global connections/influences.

I say, go see the world, you'll never know what you've missed until you go.

[was1984]

For what it's worth, I used to work at AMD and most of the architects there had PhDs from schools like UT-Austin, Michigan, Berkeley, Georgia Tech etc. Almost nobody had a degree from Stanford, and almost everyone had a PhD.

That was in Austin, so it would probably be different in the bay area.