Archive for the Science Category

Broomball

Posted in Science with tags , , , , , on January 31, 2009 by zolthanite

Not really a post today, except to mention that I’m going to be running off to play Broomball in a few hours.  So I’m going to reproduce one of my old posts (Date: Nov 17, 2007) about engineering and what it means to me.  Enjoy:

=====

Today, at 9AM PST, my Numerical Linear Algebra professor died of acute myeloid leukimia. From what I understand, he has no wife, no children, as his family was his colleagues and students. Gene Golub, for lack of any other way to express what he has achieved in his lifetime, has been one of the most influential mathematicians in the world. It is because of a seminar he gave when I was a junior that I am here, studying mathematical modeling in order to find a way for my interests to better the world in the broadest way possible. But for the non-academic community, he probably has a much more significant impact than you would ever know, unless someone told you.

Gene’s crowning achievement is a numerically stable version of singular value decomposition for matrices. This means absolutely nothing to anyone that posts on these forums, much less can read it. It’s a series of black magic operations that I’m still not privy to, but I respect it all the same. However, what I can tell you is that even in the most elementary of ways, the SVD is the most important tool that we have access to as engineers, and we don’t even realize it exists.

I am fairly certain that for gamers in particular, he may be indirectly the reason that we can have a game like Far Cry 2, which will supposedly support real-world physical models correlating the the weather (Assuming they use real formulas to do so). Even beyond that, the people that work tirelessly on computer architecture and hardware, that have to create simulated models of disturbingly complex and empirical systems based on quantum mechanics to minimize the costs associated with fabrication facilities that cost in upwards of billions of dollars, have depended on his methods as much as anyone else. In a sense, we probably owe a lot of our modern processing power to, and people like, Gene. But it extends much farther from there.

Some of us are going to be driving to P2KX. Some of us will be flying. Do we know how much work goes into designing these things? If your car is American, it’s probably a series of well-educated guesses, formulated on prior work and knowledge with a little bit of “git ‘er dun” work at some point to meet a deadline. If you’re flying a 737 out of O’Hare to Niagra, you can narrow most of that work down to people like Anthony Jameson, who spent years of his life working for Boeing on fluid dynamics models that have served as the very basis for designing almost every aircraft that leave their doors for almost the past three decades.

In a related vein, one of my classmates is currently doing some research work for an adjunct professor that works full-time at Lockheed Martin. He is currently doing analysis on the wing models for a satellite they are about to launch within the year, attempting to determine if the satellite will fly successfully or completely self-destruct on itself in orbit over the Earth. Like the air flow models, or any other physical model for that matter, what determines this is the eigenvalues of a series of partial differential equations. Typically, we only care about the principle eigenvalue, as it is the largest and dictates the overall behavior of your system.

One of the ways the principle eigenvalue can be found is by performing the SVD.

Yet the world, as a whole, will never know, or usually care who Gene Golub is. As global technology has advanced, much like gaming, we forget the producers, the programmers, the mathematicians, the very people that allow us to push the limits of our imaginations and create a way for men to pretend they are women in World of Warcraft or Final Fantasy, or use GMail to contact a long lost friend and visit them using an e-ticket purchased over Orbitz instantly with an instantly verifiable Visa through your Starbucks Wi-Fi connection.

To be an engineer, a scientist, a researcher, means to spend your entire life dedicated to your craft in a way that many people would find staggeringly oppressive. If you work in industry, you rarely get credit for the work you do outside your department, with all of the accolades granted to the company, not you. If you work in academia, your work is largely ignored by industry, letting most of your results fall by the wayside, unless you are fortunate enough to have someone from the corporate labs division read your work and make the effort of pitching it to marketing and applying the effort to get theory turned to practice. But somehow, we never seem to stop caring and loving what we do.

Gene is by far the best of these people I have ever had the pleasure of meeting, and I know the worldwide academic community as a whole will miss his presence in a way I cannot even begin to fathom. Not only a brilliant person, but a wonderful personality and a genuinely great person to be around. He is the only person I know who presented research results as if a real person who poured his heart and soul into his work was responsible for them, and not as if it was merely a result that sprang from the ether with a name attached to the side packaging. My only wish in life is that I am able to be half the professor he was during his lifetime, and inspire those around me as he has inspired countless others during his lifetime.

Gene, I will see you in class in a week and a half.