Intel Corp (INTC)

[spokeshave]

Just a few points to consider about the Barton test.

First off, it strikes me as very strange and suspicious that the tests showed zero, none, nada, in the way of improvement over the Tbred. I would be willing to say that Barton is a flop if the test showed only a 5% overall improvement, for example. But that is not the case. It would appear as though they were testing two identical units. Am I the only one who thinks something is fishy? Let's play a little game. Let's all pretend that this is a review of a new Intel processor.

Clearly, there is something wrong with the test. I can think of a few possibilities.

Near the top of the page, there are a few POST screenshots. The first one clearly says "For KT-333 DDR Chipset" in the startup screen. This screen also shows a model number 2500+. The next few shots show a similar screen that clearly says "nForce2-ST Chipset" with differing speeds and model numbers.

Nowhere in the article can I find clear indication of which chips were used on the KT-333 motherboard and which ones were used on the nForce board. Clearly, if two different platforms were used, the test is bogus.

Further in the article, they do mention the EPoX EP-8K5A2 with version 4.45 of the 4-in-1 driver. Neither the EPoX site nor the Via driver site claim that the motherboard or driver supports Barton. There is no indication of a BIOS flash being performed either.

My point is that, once again, you guys are believing that doubling the L2 cache to 512 has no impact whatsoever on a processor's performance. That simply makes no sense. It is far more believeable that the test setup was flawed.

That is why I said it is better to wait for a well designed test from a well respected source.