Although I agree with m'colleague above, as I understand it, "
Moore's Law" was never an actual 'law', as such.....and certainly never on any statute books anywhere.
It was an observation - by Gordon Moore, one of the founding members of Intel - during the mid-80's to late 90s.....that at that time, the speed of progress was roughly doubling the number of transistors on any given IC (integrated circuit) approximately every 18 months. For several years, this held true.
As time passed, those in the industry began encountering issues which made it impossible to maintain this "truth" (including the prophecy that CPUs would soon be running at speeds of 10 GHz +). This was largely due to two things; excessive heat generation, and
electromigration. This latter was first encountered at the time of the Northwood-cored Pentium 4 when vigorously overclocked via the application of higher voltages, and led to what became known as
Sudden Northwood Death Syndrome amongst the overclocking community. Different approaches were then pursued, leading to the slower-running but more capable early multi-core chips of the day.
Official thread (retired "sticky" ) at Overclockers.com:-
~~~~~~~~~~~~~~~~~~~~
Nowadays, we seem to have gone full circle. Not only are chips being produced that run at insanely high speeds, they also have dozens, if not scores of cores AND many of 'em only sip at the available power. However, performance still seems to need higher power throughputs to truly attain the expected results.
'Moore's Law' has been in decline for many years, at least going by the original premise. But even if it WAS "deleted" tomorrow, precisely ZERO would happen...
(
shrug...)
Mike.
