Intel Core 2 Duo has two independent processors running at the same clock speed in one package and share 4 MB of level 2 cache. AMD initially tried to make an argument that shared cache was in fact a disadvantage because it would enable one processor to hog all the cache to the detriment of the other when multiple applications are running simultaneously. However, Intel refuted this, in effect saying that the shared cache is a large pool that both processors can dip into as required.

In its technical specifications Intel supports this by saying that its smart memory access system is designed to keep cache full of the most used instructions, using predictive algorithms to move data from system memory into cache before applications execute instructions that need the data. This would appear to indicate that it doesn't really matter which processor an application is running on, if data looks like it's going to be used often, then Intel's system is smart enough to move it into the cache.

Aside from a more efficient and larger onboard cache memory system in the Core 2 Duo range, Intel claims improved execution speed with more instructions per clock cycle.

With all this extra processor power floating around generating heat and consuming power at a great rate of knots, Intel has implemented an interesting feature to address the problem. Intel has segmented the system so that it has the ability to power up only segments of the system as they are needed. As Intel technical Graham Tucker explained previously: "We only turn on the segments of the cache that are being accessed. We've segmented the whole 4 megabytes of cache and effectively the cache that isn't used is powered down."

The result of all this is Intel's X6800 and E6700 dual-core processors have been killing AMD's top performing dual-core FX-62 AM2 in virtually every published benchmark test. AMD's only option now is to drop the prices on its dual-core range to stop too much impact on its market share. {moscomment}