I don't think you need to be an engineer to be a good CEO, but you do need to understand the company's core products and if it's a company where engineers create those products the CEO will need to know when to defer to their experience and expertise
It's engineer, fab guy, or finance. Those are the big choices. If you look at Intel, Fab guy = Barrett, Kraznich, Finance = Otellini, Swan
It seems Fab experience isn't necessarily good for the company because there's a difference between relentless execution and future vision. They excel at the former but fail at the former(of course Kraznich failed miserably at both).
It's certainly not finance, because Otellini's mistakes are costing the company
today. So from a practical perspective it has to be engineer. When it comes to respect Pat Gelsinger has loads of that. There always exist groupthink, and he's
one of them.
One article said there has to be a balance between vision and management and his experience at VMWare means he has both of that, and the needed software perspective, which he didn't have it back at Intel.
There's absolutely no doubt in my mind big part of the drivers actually improving(and rapidly) has to do with that, indirectly by Gelsinger bringing in Lavender and reorganizing the group.