X6 varying test results

[Pasifist]

I've come over an test; http://www.overclock3d.net/reviews/cpu_mainboard/amd_1090t_six_core_review/6

Does'nt this kinda stand out from all the other tests where the X6 is the looser? I'm considering buying X6 1090T, but I'm a bit worried as the I7 930 seems to usually stomp on it. However, the results in this tests indicates quite the opposite. Do anyone have an explanation for this, or for why these tests vary so much. Enlighten me

 

[unknown_13]

Yes, in games the i7-930 beats the X6 1090T, 'cause of the better architecture and because games won't make use of the additional 2 cores. For gaming there is absolutely no need to buy the X6 10xxT CPU's. Also, the i7 CPU's are a bit of a overkill for gaming. If you're strictly gaming, get a i5-760 or a PII X4 955.

 

[Pasifist]

Yes, but have a look at the review I linked. It tells a quite different story.

 

[unknown_13]

And most of the other reviews tell the "real" story. Dunno what is the problem with those benches. (except for the 3dMark benchmark, it tends to use all available cores....)

 

[Pasifist]

Yes, but what excactly makes the results in this test vary from other results? I mean, to assume there's a problem with those benches is a quite simple way to look at it? I've read through it all and the test seems to be serious and professional review? If there clearly was a problem with the benches the review wouldn't have written that review that way?

 

[unknown_13]

Hmm..... i really don't know.......

Anyway, you said you wanna bought an X6? What will you use the PC for?

 

[Pasifist]

Mostly gaming

 

[unknown_13]

Mostly gaming

Then get either an i5-760 or a PII X4 955.

 

[ElMoIsEviL]

Yes, but have a look at the review I linked. It tells a quite different story.

The review is biased? Not trustworthy possibly?

If every other review tells one story and this single review tells another than the logical conclusion is that this review is not trustworthy. There is something wrong with either the way the testing was conducted (a mistake) or the authors are biased losers with an agenda.

Hope that answers your question.

 

[unknown_13]

The review is biased? Not trustworthy possibly?

If every other review tells one story and this single review tells another than the logical conclusion is that this review is not trustworthy. There is something wrong with either the way the testing was conducted (a mistake) or the authors are biased losers with an agenda.

Hope that answers your question.

+1

 

[Wisecracker]

Or the guy knew what he was doing, disabled Turbo, OC'ed the 1090T to 4.1GHz+, cranked the IMC/NB to 2800MHz+, and was using an ASUS Crosshair 890FX with a Noctua NH-D14.

I see no bias. What I see are 2 very similar CPUs when each are OC'ed to 4GHz+.

As far as the gaming is concerned, Dirt2 is known to favor AMD, anyway. Not sure about NFS-Shift.

And Crysis is Crysis.

To claim a review is biased and untrustworthy because you don't like the results is the last refuge of a scoundrel.

 

[ElMoIsEviL]

Or the guy knew what he was doing, disabled Turbo, OC'ed the 1090T to 4.1GHz+, cranked the IMC/NB to 2800MHz+, and was using an ASUS Crosshair 890FX with a Noctua NH-D14.

I see no bias. What I see are 2 very similar CPUs when each are OC'ed to 4GHz+.

As far as the gaming is concerned, Dirt2 is known to favor AMD, anyway. Not sure about NFS-Shift.

And Crysis is Crysis.

To claim a review is biased and untrustworthy because you don't like the results is the last refuge of a scoundrel.

Scoundrel? You ought to be banned for that.

It is only logical that if 25 various independent sites all point to one particular result while a singular other site contradicts those results... that.

A. That singular site did it wrong.
B. Biased.

It is common knowledge (not sense but knowledge) that the Nehalem Architecture excels in gaming scenarios and that the Phenom II X6 does well in multi-threaded scenarios.

 

[keithlm]

To claim a review is biased and untrustworthy because you don't like the results is the last refuge of a scoundrel.

You are wrong. The facts as erroneously presented in various forums prove it.

Apparently on some forums it is acceptable to claim that a review is biased and unacceptable if it shows AMD doing well.

However on those same forums the reverse is not true: you aren't allowed to call a pro-Intel benchmark review biased even if you can point out half a dozen things about the benchmark that were done incorrectly.

Like the NB speed as you mentioned, or perhaps using very loose timings with slower memory on AMD and tighter timings with faster memory on Intel. Or not bothering to increase the NB Voltage, getting 400Mhz less than the average max overclock and then doing a complete review with the Intel chip using a 400Mhz higher clock. (Only someone completely without a brain would do something like that.)

Or comparing the chips "at stock" and posting the base clock frequency for various benchmarks on the result charts even when you know the chip is not running at that base clock speed. That takes extreme bias. (Or complete desperation.)

It is common knowledge (not sense but knowledge) that the Nehalem Architecture excels in gaming scenarios and that the Phenom II X6 does well in multi-threaded scenarios.

What you say is commonly accepted on some forums; it is not "common knowledge".

But then we still have some less than knowledgeable people that believe Crysis is the last word in gaming benchmarks. They don't seem to understand that if Crysis had not been so pro-Intel that it would have disappeared as a popular benchmark years ago. Just like has happened with a dozen other benchmarks that didn't really show a difference between Intel and AMD. But because Crysis has anomalous results the pro-Intel people LOVE it. (Instead of throwing the results out as being statistically unacceptable.)

The same thing applies to some popular non-game benchmarks. You can show 4 similar benchmarks that compete with the questionable benchmark but since their results don't completley agree with the "popular" benchmark they are thrown out. When in fact the reverse should have been done.

 

[Wisecracker]

Scoundrel? You ought to be banned for that.

It is only logical that if 25 various independent sites all point to one particular result while a singular other site contradicts those results... that.

A. That singular site did it wrong.
B. Biased.

It is common knowledge (not sense but knowledge) that the Nehalem Architecture excels in gaming scenarios and that the Phenom II X6 does well in multi-threaded scenarios.

We all know what opinions are like.

Where is the bias which is the basis for your claim?

Where is the contradiction? Where are the outliers?

What I see is amazing consistency with other reviews across the board in the article; not only in synthetics, but apps, too.

What it comes down to is you are mad and disillusioned because he benched Crysis with 0xAA and noted "gaming tests are 1920x1200, maximum in game settings with 8xMSAA" ...

and you don't like the results ??

LOL at you