If the problem is the architecture of the elements on the printed circuit board, then I think manufacturers can fix it in a timely manner by changing the configuration to avoid such overheating, the main thing is to recognize this and find a technical solution, just of course such a start for a new generation of Nvidia cards is alarming and will make many refrain from buying Nvidia Amper cards for a while, it's good that I refused to pre-order them.
The issue as I see it was a tradeoff between better heat dissipation vs higher stable clock, One type of cap had better
heat disspation but was less stable at higher clock rates.
I don't know what went into the decision making so I won't challenge the decision one way or the other. The real problem,
IMO, is they didn't engineer the clock rates properly for the type of cap used.
This makes it more of a PR issue than anything else and I fear a PR solution. Had the boost clock rates been specified correctly
we wouldn't be having this problem. Now to save face they will be forced to meet the incorrect specs, likekly at some less
immediately visible cost, like shortened life span.