You're better off looking for specific professors who are doing research in the areas you're interested in rather than choosing specific schools.

to AnalogDE:
Thank you for your reply. I agree with you and actually I search potential professors in this criteria.
I ask you guys for advice to see if I have negelect some potential ones.

to RFICDUDE:
Thank you very  much for your advice both on motivation to pursuit a Ph.D. degree and how to select a professor.
Briefly speaking, my first goal is indeed a good opportunity with a leading IC design company. I want a Ph.D. in the following three reasons.

1. So far as I know, Ph.D.s often have greater oppotunity to be  a member in leading IC design company. Moreover they do higher-level jobs at firms.

2. I want to do some new is this area. Although firms are the place that you can do new things, schools is considered to be a better place. Even if there might be many failures before the ultimate success comes, the penalizations at school would be much cheaper.

3. This one my be a little utilitarian. I consider that , after some years in industry,  I may take an occupation in a university, just as many professors do today. In this sense, a Ph.D. degree would be necessary.

I know these might be immature. Any comments from you and other guys are appreciated.

I can only underline what RFICDUDE wrote down.

A couple of years back in my final year at university, I found myself in the position: do I want a PhD, or do I want 4 years of industry experience.

When I looked at PhD students at my university, they really looked miserable and if you would ask them like what project are you on, they refused to tell anything. From some I received some feedback, and I realized the professor was just using them for cheap labor. Not really innovative or anything. Of course they tried to make it sound special. And then there is the politics of universities. Large bureaucratic systems. Either you like it, or not.

So anyhow, I started my career in industry, and I was amazed with the knowledge there. I feel that I've learnt and seen more than that I would as a PhD. After a few years I got to tutor even a newly hired PhD. What an irony

So you have to ask yourself, what is the PhD title worth to you?

This could be bit late...but this is a famous question :-D

I think its very personal choice, but I would still like (cant control myself in getting involved in this) to make some general comment. You said ckt design, I think that means analog ckt design. Coz digital ckt design happens at a much higher level of abstraction. Anyway...

Analog Ckt design is still done best in industry, if you consider 'best' to be what eventually works without any hidden weaknesses! But as already pointed out something grand new rarely happens! Its almost in all cases some tweaking to previous designs to obtain something new. But you get to work on which 'works' in Silicon. No fuss! Since a company is a business unit, its driven by business requirement. If a company can make millions of dollar by selling same product years after year they will do that (similar to EDA industry)! But thankfully that doesnt happen, there is always competition to come up with a better chip. Now better could mean a lot of things -- cheaper per functionality, better spec, more compatible with the end-customer. The last point really determines a lot many projects to be killed, simply because the customer company doesnt want to change their board :-O
In these cases even a better spec design may not win the business!! But since some smart customers like to change, chip design companies also need to come up with better soln. So broadly in any such company there are 2 design grps -- i) product line & ii) futuristic design grp. Most of us land in the former as this is the grp that keep the money flowing for the company, where the 2nd grp mostly investigates on future trends (many say they waste the money the former earns ). For the 1st one you dont need a PhD. You need to understand circuits. You learn a lot initially to understand the design flow (if you are in digital grp, thats the heart of the whole thing), you work on real ckts, and different ckts teach you abt working principle of different ckts. This is nice! But to enter the latter grp you need to have specific knowledge on something. You will be hired mostly for some strategic reason --- in other words you better have a PhD or have good work exp on that specific topic.

One more thing, if you enter the first grp after doing a PhD, you may see a lower-degree person is mentoring you, as doing a PhD on some specific ckt doesnt make u an immediate pro in all the kinds of ckt that this grp is involved in designing.
One needs to understand simulation tools as well, I have seen good ckt designers struggle with simulation tools and mangers dont like that :-D Your qualification and enormous knowledge on something (which u gathered during ur phd) may not rescue from this day-to-day stuff.

I generally wud suggest to spend some time in product line in the beginning to understand your interest, because more often than not, we dont know what we want. I am saying this more, because the question you asked itself reflects this uncertainty. But if you are more interested in obtaining the degree then go ahead, as things get a bit tougher in later stages...and you may lose the motivation!

cheers!
Rajdeep

If you ever want to teach at the university level, stick it out and finish the PhD, otherwise industry experience is generally more valuable as a designer.