We have also checked out the performance of a similar 3.0GHz Kentsfield (Core 2 Extreme QX6850) processor in identical configuration as the Yorkfield processor described above but on an Asus P5K3 mainboard. The results in 3DMark06 were as follows: total score ¨C 11,986, CPU score ¨C 4,393.

Of course, this is hardly enough to draw any final conclusions, but 3DMark06 tests indicated that the 45nm processor ran 1% faster than the predecessor in the graphics subsystem test and is only 5.2% ahead of it in the CPU test.

So, it looks like Intel doesn¡¯t really want to show direct side by side comparison of the opponents, because the BIOS of their Bonetrail mainboard is still not completely finalized.

Intel X38 Bonetrail mainboard

Further platformance optimizations are most likely to let Yorkfield processors get farther ahead of their predecessors. Nevertheless, the advantage we have already seen these days may become a pretty good argument in favor of 45nm CPU that is positioned only as Kentsfield refresh. In other words, the new radix16 divider and 16MB L2 cache can already guarantee pretty decent performance boost even without SSE4 instructions involved.

However, it is still too early to make any final conclusions. When Intel finishes their platform optimizations we will have a better chance to see where the relative Kentsfield and Yorkfield performances are.

And in the meanwhile we can offer you some "publicly available results" provided by Intel at the forum. As the diagram shows, Yorkfield outperforming the predecessor by 7-63%.

By the way, the phenomenal result in DivX 6.6.1 encoding application can be explained by the fact that this codec already supports the new SSE4.

In fact, we encourage you to regard these results with a certain grain of skepticism, because these are no independent tests, of course. So, stay tuned for our extensive test results closer to the Intel Core 2 Extreme QX9650 launch on November 12, 2007.