This seems like a failure of algorithm design to me, in the sense of not starting with a better model that more accurately encodes the possibility of timing attacks. That being the case it appears to me to be a problem still firmly residing in the domain of computer science.
But, I'm a programmer, not a chip designer, and I have very little knowledge of this field, so I'm probably biased and not thinking about this correctly or with enough nuance.
> Computer Science likes to live in proximity to pure mathematics, but it lives between EE and mathematics
which doesn't disagree with "firmly residing in the domain of computer science" - it's merely a question of nobody having factored the right bit of EE into the right bit of math to get the right model.
I see; so you're talking about step two as described here, do I have that right?
https://en.wikipedia.org/wiki/Spectre_(security_vulnerabilit...
This seems like a failure of algorithm design to me, in the sense of not starting with a better model that more accurately encodes the possibility of timing attacks. That being the case it appears to me to be a problem still firmly residing in the domain of computer science.
But, I'm a programmer, not a chip designer, and I have very little knowledge of this field, so I'm probably biased and not thinking about this correctly or with enough nuance.