I have been asked on several occasions for my views on the on-going question, “how many analog engineers are there?”, usually as a preface to a discussion of the alleged shortage of engineers in general, and specifically analog engineers. Despite the fact that I tell inquirers that all I can offer is a SWAG (scientific wild-ass guess) response, they still insist, as if I have some special insight.
Well, I do have insight, and it's this: the question can't really be answered. As with so many other real-world queries, the answer depends on your definitions, perspective, and motive. Further, any answer you do get will have a large error band.
First, there's the basic question: Is an analog engineer someone who designs analog ICs, or is it someone who uses these ICs to develop analog subsystems and circuits? How do you count an engineer who designs an analog interface for data acquisition (analog), but also does the PC interface (digital) and the drivers (software)? What about the digital designer working with a gigahertz-range clock, who has to revise a board layout due to EMI, signal integrity, crosstalk, stray capacitance, and too-hot IC issues? How do we count design engineers who are not “analog” in the strict classical sense but are, due to the unavoidable laws of physics? And are power-supply designers considered analog engineers? Just North America, or worldwide? Where you stand on the question depends on where you sit, as is often the case.
Perhaps there's another approach: let's look at the number of EEs getting Bachelor's or Master's degrees in analog design. Again, this is misleading, since many schools don't offer analog design as a defined concentration, although some students do focus on it. Further, many engineers get involved with analog design only after they have left school and taken actual jobs.
Our inability to answer with an error of less than 30% either way (that's my error-budget assessment) to this seemingly simple question is certainly frustrating. We are used to measuring time and space with extraordinary precision, over many orders of magnitude spanning prefixes of femto (10-15 ) to peta (10+15 ) and beyond. Yet we can't get our hands around this should-be-countable, discrete question, and that's contrary to our quantifiable urges. Get used to it, that's all I can say, and that's with a mix of frustration and resignment.
Bill Schweber , Site Editor, Planet Analog