Originally posted by: Mingon
If you compare the PIII on .18micron (1ghz max) with the PIII tualatin (1.4ghz - 1.7ghz o/c) this perhaps hint to a 0.13 t'bred being able to hit 2.5ghz possibly.
1.1GHz max for the .18U PIII..... and the cD0 core was known to o/c as high as 1.2-1.3GHz reasonably well.
But their simply not comparable though, considerably more was involved in the die shrink from .18u to .13u. It was a complete re-layout of the core, with additions, added copper interconnects on a completely different architecture the the K7, not to mention brand new equipment and litoography tools specifically intended for .13u production process. One could go on about the differences, but suffice to say they were under dramatically different circumstances.
This compared to a bare dumb shrink of a core that is already long passed it's 'sweet' spot in terms of process design an clockspeed.
It's been said before but it needs re-stating... one simply cannot compare to very different cores and different processes to determine potential clock ramping.
FWIW though, I do consider the P6 core perhaps the most phenominal X86 architecture ever designed. It's life span was remarkable and almost singlehandedly pushed X86 to performance levels once though to be completely unreachable on such a 'poor out-dated' ISA.
A design that begin with so many complains about poor performance in the PPro, and was never expected to need to scale beyond 800MHz... and wasnt expected to have a maximum potential beyond 1GHz... has hit the market in volume at clockspeeds of 1.4GHz and beyond.
Facebook
Twitter