"fudging" the price/Performance Charts?

http://images.tomshardware.com/2007/02/03/graph2.jpg

Firstly, I think Tom's does a real service with their price/performance charts, it's generally good data, logically presented.

I can go to how much I want to spend, and see how the CPUs I can afford compare in that price range.

That being said, a lot of Intel CPU's simply are not on the chart which is, due to their absence, quite misleading.

Examples:

A c2d E6300 @ +/- $195 and an "estimated" Performance Index (PI) of +/- 1.55 compares well to other offerings:

E6300 - PI = +/- 1.55 - Price = $195
x2 4600+ = PI = 1.57 - Price = $213
E4300 / E4400..?

Lower down the food chain, the Netburst D9x5 chips are also missing

D945 - PI = 1.45 (same as D950) - Price = +/- $175
x2 4200+ - PI = 1.44 - Price = $169

x2 3800+ - PI = 1.35 - Price = $136
D925 - PI = 1.29 - Price = $145

Seems odd to me a huge volume chip like the E6300 has not been tested...?

This is especially "interesting" given the write up with the last set of price/performance charts..

Toms says...

Graphically, we can see that, at around the performance index levels of 1.35, 1.45, and 1.55, AMD offers the lowest price points.

http://www.tgdaily.com/2007/02/03/ [...] page2.html

Umm. that is EXACTLY where the E4300 and E6300 are very, very attractive...

What? - Tom's has never heard of an E6300? - This is a part which directly contradicts the thrust of the article....

E6300 - PI = +/- 1.55 - Price = $195
x2 4600+ = PI = 1.57 - Price = $213
E4300 / E4400..?

Thoughts???

"official" comment from Tom's...........????

They already have an E6300 that they have been testing with. Not this week but the 2 weeks ago Price/Performance mentioned that they are testing with the E6300 and E4300 and plan to add it to the chart. I have already send Email to the editor. That was back in December. They informed me that they would be adding the E6300 soon to the charts. I thought soon was going to be in 2 or 3 weeks. It has been much longer than that.

Maybe they are waiting for a new stepping of the Conroe to be released. From the looks of things that would be something beyond the E0 internal stepping. For all we know Intel could be up to G or H stepping.

The problem with setting an E6400's multi down to 7 is that when tested in a 965 chipset, it'll increase the NB fsb 1.167:1. There's a quirk in that it advances the NB fsb speed if you're not at the default multiplier. More info on this is in the link in my SIG.

It's a good thought to slow down an E6400, but it has it's caveats.