As described in the last entry, the International Solid State Circuits Conference (ISSCC) is the largest and most important electronics conference in the world. That entry listed which countries and US states contributed the most papers to it. How about organizations? Which companies, labs, and universities do the most significant electronics research? I did a similar tally for the conferences at decade intervals starting in 1960. For brevity’s sake I dropped organizations that had only contributed one paper in all that time. Some organizations have spun off other operations in this period, and I’ve tried to list the ones I know about together.
Let’s start with the companies that have contributed papers:
|40||Bell Labs, Lucent||6||13||6||3||12|
|4||Bell Northern, Nortel||1||1||2|
The leaders here, IBM and Bell Labs, are not surprising. They have both been major innovators since the 19th century. Notice, though, that Bell Labs had nothing in 2010. They’re now part of the French company Alcatel, and are still somewhat active, but only in far more applied areas. This is part of the ongoing collapse of corporate research in the United States. IBM is the only major company that still does basic science. Notice how GE faded long ago, HP and Motorola are on their way out, and RCA is gone altogether.
Also notice that Philips is the only European company in the top 10. Electronics appears to be the domain of the US, Japan, and recently, Korea.
How about research labs and universities?
|7||Delft U (Netherlands)||1||6|
|7||UC Los Angeles||1||4||2|
|6||National Taiwan U||6|
|4||ETH Zurich (Switzerland)||2||2|
|4||Hong Kong U||1||3|
|4||U Leuven (Belgium)||1||1||1||1|
|4||U Pavia (Italy)||4|
|4||UC San Diego||2||2|
|3||Fraunhofer Institute (Germany)||1||2|
|3||Keio U (Japan)||3|
|2||Helsinki U (Finland)||2|
|2||Oregon State U||2|
|2||Shizuoka U (Japan)||2|
|2||Tohuku U (Japan)||1||1|
|2||U Tokyo (Japan)||2|
|2||Yonsei U (Korea)||2|
Again, no surprise at the leaders: MIT, Stanford, and the University of California at Berkeley. However, the UC system as a whole has 25 papers, twice as many as any other. The different campuses do have separate identities, though, so I’ve split them up here.
Of more interest are the institutions down the list. IMEC is the first non-university shown. It’s a huge government lab (it employs 1750 people) in Belgium devoted to micro-electronics, and it’s become a real center for it in Europe. The University of Leuven is near it, and has been doing important research for a long time. Because of Philips I knew that the Dutch had been doing big work in electronics (and now Delft University is important), but who knew that the Belgians were so active?
The only other non-unversity is the Fraunhofer Institute. It’s a vast enterprise at this point, employing 18,000 people in 60 centers, with a budget of ~$2B. About 7% of that comes from its patents on MP3. It mainly does contract research, but gets 30% of its budget from the German government.
In the last decade there has been a lot of work from Asia outside of Japan, but also notice how work is spreading in the US to various state universities. As the conference has grown, and as electronics has spread, we’re seeing contributions from more and more parts of the world.
Hi John – this is interesting; my curiosity extends to how you did this. If you could get the data for this conference formatted in a SQL database, it would be pretty easy – but where is such data available? Or were you really shuffling printouts of the conference indexes?
Sadly, this info is not in a database. The Solid State Circuits Society does publish a DVD with the ISSCC papers since 1955 and the entire run of its journal, but it only indexes the paper titles and the authors, not their organizations or geographic areas. Those were shown on the scans of the papers themselves.
I wonder if that was deliberate. Perhaps the Society wanted to concentrate more on people than on organizations or countries. Country info, after all, could prompt invidious comparisons between nations. That’s never caused trouble in the past!
SRI is a nonprofit so it belongs in your second table (not that it’s contributed much either way). Funny, because it does do a lot of physical science — my neighbors are materials scientists who have lots of shredded aircraft parts on display. Hm, but it now includes Sarnoff which is probably responsible for most of the RCA papers.
I wonder how a similar study for software would go? There are a few companies with serious research labs (MIcrosoft and Google, obviously, and IBM/AT&T and a few others).
I just stumbled on this which partly answers my question: http://jeffhuang.com/best_paper_awards.html