If you're an experienced analog designer, today's laws of supply and demand mean that you're a hot property. On the other hand, if you're a system designer, you'd probably like to get some well-seasoned analog designers in your group. If you're an engineering student, well, that's something to think about. But becoming an analog engineer isn't just a matter of signing up for the right courses.
I got an array of responses when I asked engineers at various companies how they cultivate analog engineering talent. Some frankly admitted to cherry-picking engineers who already had a number of years at noted analog companies. Others have close relationships with the graduate schools at first-class engineering schools. Still others acknowledged an obligation to work with undergraduates in internships, or at least to bring in recent graduates, integrate them into teams, give them real work to do, and let them continue to learn and grow through hands-on experience.
In my opinion, engineering students should select their school carefully. Then they should talk to the placement office and professors to find out about analog interning opportunities. If no active programs exist, they should think about changing schools. The odds don't favor your sudden transformation into an analog engineer after graduation.
When I asked companies what working engineers could do to move to the analog side after a few years in the digital field, my subjects tended to lose eye-contact with me. The most hopeful response I got was a suggestion that a digital engineer might move into applications engineering and try to interface with customers whose design needs spanned the digital/analog divide.
If analog engineers are in short supply in North America and Europe, the need is even more acute in Asia. One vice president at a large U.S. semiconductor company who was born and educated in India recently confided that schools in India and China have brilliant faculty and bright and eager students. Yet they rarely have the economic resources for the latest lab equipment and design tools with which to actually design, fab, and evaluate real chips and circuits.
So at least for the immediate future, opportunities remain in engineering, albeit different opportunities than we thought about a few years ago. The challenge will be in making them pay off over the long haul.