Most engineering undergraduates end their post-secondary education with an EE, or local equivalent. Many Physics graduate students were engineering physics undergraduates, a mix of engineering and physics. They spent four years learning about their specialty but as with us, all the education in the world cannot replace hands-on in the real world. Those students who go on to a Masters program or are doctoral candidates learn to apply all that knowledge in the research labs. This is their trial by fire. What is a good safety margin? Is Absolute Maximum to be taken literally? How extensively should simulations be? Does 1KV do more than curl your hair? Can you touch a resistor dissipating 10W in free air? Sound familiar?
These post-graduate students are the cream of the crop from the undergraduates. Define a task, train a bit as required and then check in occasionally. New fields are learned fast. There is an astounding ability for graduates with even little programming experience to pick up embedded Basic and integrate it with an HDL or GUI language. Hardware takes a little bit longer than software to catch onto, especially analog. The problem is the orders of magnitude and what is possible. One of my first projects at the university was to switch currents two orders of magnitude above my experience at rates three orders faster! ‘All the numbers work, so what’s the problem?’ I was asked.
Most physics research is pure research, too esoteric to be commercial. Companies struggle to be the first to produce products using the research but the true results are years or even decades away. From Quantum Optics has come quantum encryption, a scheme considered inherently uncrackable and unsinkable. Probably true. A holy grail, but not the target of the research, is quantum computing. It would be to the computer as the computer is to the abacus. Probably true too. Industry realizes all this and is waiting.