Whenever I see another report about an engineering shortage, I don't know whether to ignore it or get furious. The latest was the story “Analog design expertise is rare, valuable” (www.eetimes.com/showArticle.jhtml;?articleID=208400206) which discussed the alleged shortage of analog designers.
My view is this: it's impossible to measure such a shortage, and it's insulting to this audience to speak of one so easily. Salary is the only half-way meaningful metric, yet it is hard to uncover and includes a lot of confusing and overlapping job titles. Without meaningful data, a claim of shortage is just the statement of a potential employer saying, in effect, “I couldn't get the people I wanted with the exact skills I wanted, at the price I wanted.” Hey, life's not fair, what can I say?
There is no lack of engineering talent to design analog ICs. I see lots of high-performance, all-analog or mixed-signal ICs announced every week; if there is a shortage, who is doing those designs? The top-tier analog vendors use the obvious and effective approach: they take people with potential and train them on their analog-IC design subtleties, including interfacing, process, and mixed-signal idiosyncrasies.
Ironically, I think there really is a very different kind of analog shortage: the declining number of engineers who actually can connect or wire-up an analog circuit and do real-world interfacing. Today's ICs implement so much of the system-level circuitry that a large part of the “analog engineering” of a product is related to external connectivity and interfacing. Getting additional IC designers won't help this problem, since the design constraints, tool, and resources for analog IC designs are completely different than they are for doing analog, system-level design on the PC board.
The real problem is that today's products, which use highly integrated, high-performance analog ICs surrounded by nearly invisible support components (primarily passives) don't leave a lot of degrees of freedom in circuit configuration and experimentation, Try changing an input resistor on a real PC board, or patching in a small RC filter somewhat, and you'll see that IC design is not the real problem.
Instead, it's the impracticality of doing analog circuit-level experimentation and investigation, which discourages real analog-circuit expertise. It's not an IC designer shortage that worries me at all. It's the withering away of circuit-level know-how, due to physical inaccessibility, that I worry about.
– x x x-