Good time to be a physicist!

[Bernard Murphy]

Not the way you think in the semiconductor world. This is more in Big Data, Deep Learning and all that other stuff that's trending. Turns out a lot of hot companies are recruiting physicists for their math backgrounds coupled to modeling experimental data. And the pay is a lot better, as I can personally attest.

Google, Facebook, and the Rest of the Silicon Valley Crowd Need More Physicists | WIRED

 

[Maxim]

Being a physicist myself (and later on - device physicist and now EDA physicist), I can tell that the main differentiation of a physicist (from, say, a mathematician or a computer scientist) is in a special, deep, intimate insight into and grasp of the problem, enabling to through away unnecessary (complicating) details, but retaining important effects and relationships, and reducing the problem to the simplest form or model. This model should be accurate (enough), practical, computationally efficient, verifiable, and useful. That's a good starting point for creating a commercially successful EDA tool

Regarding "semiconductor world"...
Modern day "semiconductor world" (at least a big part of it - circuit design / layout / fabless) has very little to do with "semiconductors" (meaning - electrons and holes, band diagrams, ionization energy, carrier distribution function, doping profiles, etc.), but rather - with various "wrappers" around that - PDKs, EDA tools, design rules, environments, tools integration, etc.