Note: By startup mentality, I mean colleges where the school and a decent amount of the student body is accepting of the entrepreneurial mindset and fosters the Silicon Valley culture (even if it isn't in Silicon Valley).
At least five professors in our electrical engineering department have companies.
The business school has successful entrepreneurs that can meet with students (in any major) trying to start companies.
There's a whole institute at the university called MTECH which has successful entrepreneurs and investors who help startups get off the ground. They have office hours once a month so if you impress them they can help introduce you to the right people.
I attend Brock University in Southern Ontario. There's not much startup mentality here but it's increasing. A business incubator was recently founded [1]. From the tech side, I'm usually the one encouraging the startup mentality (with a few friends). For example, we organized the school's first hackathon last semester. Some of our graduates have created startups, but I'm not aware of any tech startups. Sometimes startups do recruiting here but not often.
Some of the schools I would recommend in Canada are University of Waterloo and University of Toronto. Both of those have a strong startup mentality. Incubators at both schools, lots of successful tech startups from both, and their locations are great for recruiting talent.
http://en.wikipedia.org/wiki/Babson_College
http://en.wikipedia.org/wiki/Olin_College
I also see a lot of students from Northeastern University at startup events here, and hear nothing but good things about their Co-Op program, where internships with companies (including startups) are part of your degree program:
http://www.inc.com/magazine/20110301/how-the-university-of-u...
National labs are great at coming up with new technologies. They are absolutely horrid at commercializing them, which is caused mostly by heavy-handed government policies.
Thanks.
Some of these problems have commercial applications, some are strictly of value to the government, and some are so far out there that no one's really sure where they'll end up (how again was ARPAnet supposed to be used for defense?).
Two different points:
The problem with commercializing government technology is that you need an entrepreneur to push it. The labs do make overtures to allow their well-paid and securely-employed scientists to leave the nest, but the scientists seldom want to stop being scientists, give up a comfortable and tenured six-figure income, to network and sell when if they wanted to network and sell they wouldn't have gone for the Ph.D. in the first place.
If I'm in a metro area where I can, without moving, find a job making a comfortable living doing what I like after my startup fails, I'm much more likely to take that plunge. In "company towns" like Los Alamos or Oak Ridge, there really isn't any chance of that.
This leaves a relay baton drop between scientists who don't sell and would-be entrepreneurs with nothing worth selling.
No problem, the labs can license their stuff out. Problem, the terms of these licenses usually carve out a pretty generous chunk for the lab that spawned the technology, and the implicit terms include delays and risk of expropriation.
The terms of the lab carveout are often pre-ordained, based on revenues instead of margins, and don't take into account whether or not a business model can support them (because it has to be fair and defensible, and how do you know ahead of time what's fair?).
This leaves a would-be licensor to choose between many worthy projects, and as you want to be close both to your inventor and possible funding sources, this creates a bias toward licensing big-city lab technologies like those found at ANL, LLNL, LBL, NREL, and SNL-Livermore.