Discussion in 'OT Technology' started by NPT, Oct 1, 2006.
Is that what Intel was saying?
I have been figuring that it is Intel looking for a new benchmark to sell processors buy. I don't think Gigahertz is going to sell too many more processors and Intel knows it.
That and being able to do massive parrellel processing on a single chip does have a certain appeal. You have no idea how much companies are willing to spend to just do parrellel processing on just like 8 chips or 16.
This is the normal progression for a new consumer processor.
First they send out a relatively "unfinished" heat-dumping processor > Then they streamline the basic design for less voltage with the same calculating speed > After this comes the "enthusiast" processor of similar models, based on the same design, which are able to clock speeds much higher than advertised and are bought at large by the consumer PC-modification community
Intel, in fact, knows the opposite. The P4's design goal was to maximize clock frequency. The idea was that they could improve performance by simply increasing the clock frequency and this would give them a marketing advantage over their competitors. Indeed, many people (many of whom fancied themselves as computer enthusiasts) were confused about the meaning of the processor's clock speed (and still continue to be confused, today), which prompted AMD's new (then, although not new, anymore) naming convention to try to combat this stigma.
However, the plan backfired when Intel found that they could not increase the P4's clock frequency much beyond 3GHz, instead of the projected 10GHz that they were planning. Hence, the radical new shift to "Performance per Watt" (remember, Intel is big on marketing) and multiple cores, both of which have proven to be easier engineering tasks...
I assure you, if anyone can make single core processors faster, for the same price, than multiple core processors, they would be doing it!