I’m a retired Sr Staff System Design Engineer from Marvell Semiconductor and was also their resident expert in High Speed Digital Design, though I held no advanced degrees beyond a high school degree.
I have an undiagnosed 2e condition such that I have a gift for logic with a disfunctional short term memory which effected my ability to learn through lecture or textbooks. However, I can be a quick study if allowed to ask questions in real time. Therefore, I had aquired nearly all of my expertise on any subject matter as an autodidact.
While still trying to work my way through college, an untimely work injury just prior to the 87 recession had forced me to give up on getting a degree when the state of CA decided against financial assistance towards finishing my degree and instead only offered training in Printed Circuit Board design, which is how I ended up as a PCB Designer in 88.
I once designed a PCB for a Stanford EE professor who liked to sit in on the design process. This gave me a chance to learn electrical engineering as I designed his PCB which I realized would only take a couple of days.
He started at the beginning of his course material and methodically covered everything sans the math.. Halfway through the second day, he stopped. When I asked why, he said he was done and had covered all three semesters worth material saying that I had a better understanding of the material than any of his graduate students..
Fortunately, I never had to take a test to prove what I'd learned which is problematic for me, but the lesson proved to be invaluable in helping me to create the proper models of electromagnetic wave behavior in the PCB environment so I could create the methods, strategies and deaign techniques that could deal with the paradigm shift in technology when wavelengths hit the PCB scale.
I had a complete solution in the early nineties though my work wasnt widely accepted until over a decade later when the industry had finally stalled out completely with the rollout of Gigabit Ethernet. Not a single design with a gigabit port could pass without failures in EMI and SI issues.
Marvell Semiconductor, had hired me a few years before that time after I predicted the anomalies they would experience from the test socket they were using to bench test their Fast Ethernet products. Marvell actually entered the Gigabit market two years late but it didn'nt matter because no boards could pass EMI or SI tests. My designs were the only ones in the world that workeed and they all worked without a single issue.
Marvell captured 80% of the market and I became the recognized world authority on High Speed Digital Design. I still held no degree, but as the only person in the world that could design working PCBs for gigabit Ehternet it wasn’t like the industry keep ignoring my understanding of the effects of EM waves had in the PCB domain..
It wasn't just a fluke, and I proved it through hundreds of design reviews that came to me from all over the world for designs that were failing and the designers or engineers working on them could not figure out how to get them to work.
For all future reviews for customers I encouraged them to provide the designs only without any lab results that revieled the anomalies causing the design to fail.
This allowed me to predict and describe the failures that they would have seen in the lab and then they would be more likely to listen to my explainations as to why their designs were failing and to follow my instructions for how to resolve their problems through simple instructions on how to fix their designs.
Out of the hundreds of design reviews, to my knowledge, 100% of them were resolved on the very next revision.