Ding's Home

For all EE programs ranking, the top 4 are always Stanford, MIT, UIUC and UC Berkeley (not necessarily in that order). These are all very very good schools. Could you compare Stanford and UCB more deeply?

To compare Stanford and Berkeley's EE programs in depth, I need to put them in the context of the whole microelectronics history and in that case I have to include "the other two giants", MIT and UIUC, in order to get a meaningful comparison.

In 1947-48, three Bell Lab scientists, Bardeen and Brattain and Shockley invented the "Transistor" (they got Nobel Prize in Physics for that invention) and that is the beginning of a new EE era. Before that, the "electrical engineering" means "EM waves" or "electric power". Now people are talking about "semiconductors", "IC", "CAD", "wireless communications" and these are all start from the invention of the transistor. Two of the three genius continued to affect the whole EE world in the rest of their life.

Bardeen is Princeton's PhD in Physics. After he left Bell Labs, he went to UIUC and started two new programs there: one is the superconductor program in Physics dept, the other is the semiconductor program in EE dept. Bardeen got his second Nobel Prize in Physics for his superconductor theory (BCS theory) and he was still the only person who got Nobel Prize in Physics twice. Today, the semiconductor program he left is still the best part of UIUC's EE program and is evaluated by semiconductor industry as No.1 in US.

Shockley was graduated from MIT. He was invited by Terman (then the Dean of School of Engineering at Stanford) and went to Stanford eventually. He didn't join Stanford EE dept or establish any academic programs at that time. Instead, he did something much more influential: he started a new semiconductor company called "Shockley Semiconductors". Shockley was not a good manager and the company failed. However, the people he hired eventually started another very successful company: Intel. The founders of Intel were Noyce (from MIT), Moore (from CalTech) and later they hired another guy Grove (from Berkeley).

Noyce was the co-inventor of "integrated circuits (IC)" (the other inventor, Kilby, was from UIUC and started Texas Instrument's semiconductor business at that time). Moore was the author of "Moore's Law" which still governs the whole IC industry. Grove was Berkeley PhD in Chemistry and worked as a professor at Berkeley's EE dept for a while. He is now Intel's CEO and Time magazine "Men of the Year in 1997". Because of his Berkeley root, Intel provides Berkeley a lot of research opportunities.

With these historical backgrounds, we can talk more about these EE programs:

1. Stanford's Terman (graduated from MIT) was a man with real "vision". He not only encouraged two of his Stanford students, Hewlett and Packard, to start up their own company- HP, but also invited Shockely back to Stanford and that's why we have a "Silicon Valley" there. His vision also affected Stanford's basic philosophy: to sell the technology/ knowledge, so don't blame Stanford's "school-shop" attitude. Before Terman, Stanford was only a mediocre private university; nobody would compare it to Berkeley at that time. Now Stanford is getting stronger and stronger-at Berkeley's cost.

2. MIT is losing the game to Stanford. Historically, MIT was famous for the invention of RADAR and therefore MIT was the best for all radar-related technology (or defense-related technology): EM waves, communication and signal processing. However, MIT was too rigid to change their focus to microelectronics (or civil-related technology). They didn't ask Shockley (their own alumnus) back and they didn't follow Terman's (their own alumnus) plan for Stanford to provide an environment for new high-tech companies. I think that's the biggest mistake they've made. Many years later, MIT's microelectronics program is still weak and the State of Massachusetts wanted to copy "Silicon Valley" (they called it "Route 128") but failed.

3. UIUC is to Stanford is just like Bardeen is to Shockley. People know that UIUC really can do research and that's why UIUC has research funding only next to MIT. However, UIUC doesn't do a good job to "sell" their technology as Stanford does. The latest example is MOSAIC, the mother of all WWW browsers. It was created by Andressen, an UIUC student, but he didn't know how useful it was until a formal Stanford professor Clark (PhD from U of Utah) asked him to co-found a new company - Netscape. UIUC was not happy with that at first because they thought a school should not sell its knowledge. That's just a typical example how conservative UIUC is. In any engineering rankings, UIUC's graduates usually can top Stanford and Berkeley's as "practicing engineers" (3 out of 5 of Intel's fab manager are from UIUC), but that kind of conservative philosophy does restrict their vision as real leaders.

4. Berkeley was the first school claimed that "engineering" was not only "techniques" and should be treated seriously/academically (UIUC and Stanford began to offer engineering PhD degrees 30 years later than Berkeley did). However, among the top four, Berkeley is the weakest in terms of research funding. Berkeley is good at using limited resources to create influential results. Their facility cannot let them make hard-core research, but they can make their results accessible to the public and to form a new standard. SPICE is just a typical example. Today no circuit designer can live without CAD tools and SPICE is the mother of all circuit CAD tools. Berkeley's circuit group and MOS device modeling group are still the best in the world.

As an applicant, if you can get an EE admission from any one of these four, you must be good enough. However, if you get more than two admissions of these four schools (usually they're Stanford plus one of the other three), you need to think about the following things before you make decisions:

1. Resource for Engineering students:
"Engineering" is EVERYTHING in MIT...
"Engineering" is among the most important things at UIUC...
"Engineering" is the origin of everything at Stanford...
But "Engineering" is nothing particular at Berkeley. Of course part of the reason is Berkeley is excellent in everything.

2. Engineering alumni distribution:
Berkeley has more alumni working in Bay Area/Silicon Valley than the other three; however, most of them might just work for Stanford graduates. Compared to Stanford and Berkeley, MIT and UIUC are "minorities" in Silicon Valley. Last year MIT and UIUC just had a "joint" EE alumni reception in Silicon Valley. That would make the "total" number of alumni bigger.

3. Survival rate for PhD degrees:
Berkeley and UIUC have higher PhD survival rate than MIT while Stanford has the lowest among the four. Yes, it's because of Stanford's tough qual-exam.

4. The "big trend" for EE programs:
15 years ago, the Gourman/NRC ranking was [1.MIT, 2.Berkeley, 3.Stanford and 4.UIUC]. Now the Gourman/NRC ranking is [1.Stanford, 2.MIT, 3.UIUC and 4.Berkeley]. As you can see, Stanford and UIUC are rising while Berkeley and MIT are falling.

To me, MIT is like a "soldier"; Stanford is like a "businessman"; Berkeley is like a "politician" and UIUC is nothing but a "worker". Time changes. We are now living in a world without "war", so "soldiers" are not important as before. We are now living in a world full of "compromising", so "politicians" can not arouse peoples' passion anymore. We are now living in a world which "money" defines everything, so we need "businessmen" together with "workers" to make products work and profitable. That's the big tread I'm thinking.

PS. Since "in-depth" analysis was asked, I talked a lot (maybe too much) based on my own observation. Most of the above points are facts and some of them only reflect my personal value which may not be suitable for everyone.

From the MITBBS