Moore's Law did not "break"
Only those who misunderstand what it means and think it meant "CPUs get twice as fast every couple years". What Moore actually said that was that the density of transistors would double every couple years (originally it was 18 months, but he adjusted it in the 70s)
For quite a while doubling the number of transistors, made possible by shrinking transistor dimensions, allowed a combination of faster clock rates and more complex CPUs that did significantly increase CPU performance every couple years (not quite doubling, but a big jump) The problem is that as they did this the power demand for CPUs kept increasing, and they reached the point where people didn't want noisy fans and energy wasting PCs. For most people, current performance (hell, decade ago performance) is "good enough" for their needs, and it isn't worth it for Intel to design new cores that suck down 500 watts with exotic water cooling solutions for the small segment of people who might be interested in them.
Now Moore's Law is used to increase density in ways that reduce power usage (i.e. larger low power transistors, gating, etc.) or increase cores/cache since added complexity wastes too much power and has hit a point of diminishing returns as well.
Would you really buy a new PC just because one came out that was twice as fast? If so, you're in a single digit minority. Most people don't have any need for a faster CPU - and anyway would realize FAR bigger performance gains by switching from a PC using a hard drive to one with a SSD. Many of those who do want a "twice as fast" PC are gamers, but consoles have sucked a lot of the life out of PC gaming since many titles are designed to run on both consoles and PCs, so couldn't make much use of additional performance anyway. So their desire for faster PCs is often based on hopefulness that faster PCs would mean cooler games designed for them, when in reality most games would still be designed with a console's limitations in mind.
Biting the hand that feeds IT
