Why is it that Tom's Hardware is comparing apples and oranges with the Athlon XP Tbred and Intel P4. In theory you should be able to compare the two processors head to head. However in the real world you can't. So you build systems to test them. Here is where my objection comes in. Everyone and their grandmother knows that Chipzilla (Intel) screwed over any other type of memory architecture except for Rambus RIMM.
Which are expensive, proprietary and soon to be left by the wayside by DDR. So why are you testing high high bandwith 800 and 1066 RIMMs against not so high bandwith 2700 DDR? Wouldn't a more appropriate test for consumers and games be a direct head to head comparison of two systems using DDR or even SDRAM. Wouldn't we then see result more inline with what we have come to expect out of AMD? Personally I have a huge amount of input and say as to what my company buys. I have avoided Rambus RIMMs like the plague since their inception. Lets see a more realistic comparison between the two cpus. Not one where intel has the advantage going in. Which would you drag race with a ferrai using unleaded or a funny car using nitrometh?
The P4 was designed to be used with Rambus, while the Athlon was not. So, if you are compairing the top performing systems, which is what THG is doing, you need to use RDRAM for the P4.
The only way to isolate the CPUs would be if there was a chipset that could use both chips, and compair them on that. That isn't possable however, so they use the best that each platform has to offer.
An analogous way to relate what you are compaining about, would be to complain about NASCAR being unfair because there are several different model cars instead of the same car using simply different engines.
"Search your feelings you know it to be true, I am your... twin sister" - Darth Vader
You are about to answer a thread that has been inactive for more than 6 months. If you still wish to proceed, please ensure that your posting is original and does not duplicate or overlap any prior responses to this thread.