x86-Secret has just published an article on the PAT technology that comes to confirm all the information heard these last days. On the contrary of what Intel is stating, PAT is present on ALL i865 and i875 chipsets and is moreover activated by default on these 2 chipsets for all modes less than FSB800/DDR400. However, the i875 is the only one to keep PAT activated in FSB800/DDR400 mode due to a better chip quality. Manufacturers like AsusTek, followed by others these last weeks, have found a workaround for the i865 chipset by making it believe that the CPU is at a 533 FSB setting, even when it's at FSB800. So what is PAT really? Marketing or technical argument? For the time being, x86-Secret thinks that the first statement might be true even if Intel, during the chipset fabrication, did not think it would have so much good yields on its i865/i875 chipsets. So for the moment, before Intel changes the revision of its 865 chipsets, these chips are very well positionned in front of the 875 ones, as the 30$ increase in price can't be explained.