CPU overclockers hit a ceiling, and it happened about 17 years ago

[Manoa]

Yhe it the true Rocky, I have 4790K and since I buyed it new in 2014 still no problems, but like you I have it on distilled loop too and some reasonable OC (44/45). and what you say about the power curve on the new processors I think they made a left turn and too highly engineered modern processors to run on 90 C, so I think what you did with voltage you probably tried to keep it around 60 C and because the processors are not designed to run like that they loosed too mutch performance
if I will be buying a new processor I should remember I am going to loose many mhz if I try to keep them in reasonable temperatures

 

[fadingfool]

I think 'always' is obviously false. The issue with the motherboard partners was, clearly, a failure of the motherboard partners, not the chip; but regardless of that, it doesn't support the 'always' claim.

It's all a matter of degree. The lifecycle of an enthusiast's use of a CPU is limited more by the desire for the shiny object ahead, rather than the CPU 'wearing out'. Non-enthusiasts don't overclock.

A but of hyperbole but in general heat kills chips (especially heat cycling) and in most cases with too identical chips the one that is not frequently cycled up and down the clocks will last longer.

 

[bviktor]

Clock by itself is a worthless number without IPC. So yes, we've hit a ceiling on clock, and thus we're improving IPC instead. Something this article didn't mention once.

And that's why it's possible for a Pentium 4 from 2008 to have the same clock as a Ryzen 5900X, yet the latter be 110 times (!) faster.