General > General Technical Chat
Doctoral studies - perspective
Slh:
Personally I chose power electronics because it gave me a chance to play with magnetics and analogue. And when I get bored of those I can do a little basic digital to relax :). As a bonus it appears that everyone else also thinks that it's either hard or not interesting so there's plenty of jobs out there. Shame I'm actually hiring...
There's a lot of research going on in power electronics. SiC appears to be taking over in the stuff I work in and GaN is something for the future. Lots of new topologies, some of them even useful...
Power is required for everything, even in all of the new buzzword fields like AI. Smaller, more efficient, cheaper - the usual challenges.
RoGeorge:
--- Quote from: Rinnake on June 26, 2022, 03:11:41 pm ---We are beginning to think that those who would doing doctoral studies in cybernetics would be much more in demand nowadays.
--- End quote ---
Fake worries. The actual specialization or PhD theme is close to irrelevant. You wouldn't expect whatever is hot right now to still be the fashion in 10, or 20, or 30 years from now, isn't it? The world is keep changing. By the time you finished your PhD, the theme will be halfway obsolete.
What really matters is the process, and also the title. These will stay with you for the rest of your life. Other than that, yes, the grass seems always greener in the neighbor's yard. Right now, power electronics is at all times high demands because of the trend to switch to solar energy, electric cars and so on. Go get your PhD no matter the subject.
rfclown:
It's very interesting that you are two brothers pondering the same journey. One thing that I'm sure everyone responding here would admit... our journey wasn't completely planned. One never knows how circumstances and opportunities will arise. My journey started with a fascination with electronics as a youngster. I looked at circuit boards and wondered, "How do they know what parts to put together? I want to learn how to do that." When I was told that the profession was called Electrical Engineering, my course was set (about 10 years old). I was good at academics... if I cared; and the only things I cared about were math and science. It wasn't until I got to college that I really excelled, because I was finally getting to my goal. I loved my time in college. When I finished my B.E.E. in 1983 I wasn't in a hurry to leave. I was encouraged by my university to continue, and received a research fellowship which more than paid all my expenses. Here's were circumstances come in. I got engaged. I needed to be a provider, so after getting my M.S. I took a job with Motorola. Great job, but after time, I wanted to be back in research (like during my graduate work). I was looking to go back to school (which would have been REALLY hard since I had two kids and one on the way) and had inquired at two schools... and then a research position opened up at Motorola that I was able to get. That turned out to be a dream position (George Fisher CEO at the time), until former CEOs trashed the company (Ed Zander...) and decided that research wasn't needed, and they'd be followers instead of leaders. I've been very fortunate that I had a fantastic education (Auburn University) and the privilege of working for Motorola from 1986-2009 were I learned from many great engineers. I've been able to excel in 7 jobs since then. Do I wish I had the PhD at this point? (skip the Moto research opportunity and go back to school?) It's so hard to say. You have decide on opportunities as they arise.
Just a rambling from my experience. That said, if you think you might want a PhD, and you are in a position to get one... you most likely will not practically have to opportunity in the future to do so. I'd say grab the ring and do it.
Nothing like getting a PhD... but I decided early on go get a Profession Engineer license. It was not needed in the job I had, but it was going to be MUCH easier to do when my academics were fresh in my mind, so I did it. Has it benefited me in my profession so far? Not one bit. Do I still keep up the continuing education requirements and pay the fee every two years to keep it current... yes. Because I don't know what opportunity might arise tomorrow.
jeremy:
The majority of money I have earned in the last 5yrs has nothing to do with my PhD specialisation/topic, it is more around electronics in general. Pretty much the only people that matters for is if you stay in academia, or you have a very, very specific job in mind (ie you want to research neural networks at Google). But the latter is a very risky choice, as you are betting the farm that you can actually get that job and that your skills will not become obsolete overnight. If/when the next AI winter comes around I am not quite sure what all of the AI folks will transition into, as it seems the skills are so domain specific that they will be not really all that transferable. I am told that this was a problem for a lot of people in genetic sequencing when automated PCR machines hit the market.
rstofer:
I would think there is quite a similarity between the math required for AI and that used for data mining or optimization of cost functions in a more general, and business related, sense. The branch of math called 'big data', for lack of a better term, is becoming increasingly important at every level of business. Everybody wants to collect data and they need math types (particularly statistics) to do something with it.
Neural networks, despite the heady title, are really just another way to find a minimum of some cost function. It's just that the networks take so many inputs (tens of thousands, perhaps) and, at best, find only a local minimum. I expect that finding the true global minimum is the holy grail of AI and is still quite elusive. Where else could you say "Jacobian Matrix" and not get a bunch of blank stares?
My interest in optimization was always related to scheduling with constrained resources and everything was done in Fortran. Those were the real 'good old days' circa 1970.
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version