Overclockworkbench tests Phenom II vs i7 940 and QX9770

[informal]

Hey people, overclockworkbench seems to just broke the NDA(if they ever signed one that is),and tested Phenom II 940 vs QX9770 vs Core i7 940:
http://my.ocworkbench.com/2008/asroc..._X4-info-1.htm

Site is opening slow atm,but Phenom II results seem pretty good,considering the fact both intel QC chips ran at higher clock(QX9770 is 3.2Ghz CPU while Core i7 mostly turbos up to 3066Mhz)

edit: Cinebench 10 32bit scores seem to be way low for their Phenom II system,900-1000pts lower.In 32bit Phenom II @3Ghz should score ~10500 and not ~9600pts... I can't say what's going on there,but their results are hindered in some way in this test(only?).
Also the OCing page made me laugh since they seem to up the NB volts and leave it NB clock at default?? Maybe they don't know that HyperTransport and NB are not the one and the same thing... Note the 3Dmark06 CPU score is 14% lower than the same clocked 3.8Ghz Phenom II we already saw in Shanghai review thread(5700pts @ 3.8Ghz).These numbers really are weird,be it their BIOS version or OS issues.

http://my.ocworkbench.com/2008/asroc...locking-10.htm

The concept of overclocking the AM2 is applicable to the Deneb core too. Important factors like North bridge voltage, Hypertransport speed, CPU CLK, multilpier, vDIMM are important elements that will result in a successful overclock.

In our tests, the Phenom II X4 940 (3GHz) is overclockable to 4GHz by AIR with the use of the AMD overclocking tool or ASRock OC Tuner, the stable condition where most benchmarks will complete is 3.8GHz. In fact, a 800MHz overclock can be achieve using two ways. One, is by using a higher multiplier x 200 (e.g 200x19). Alternatively, we can run it at 15 x 253MHz. Do take note that if CPU CLK is raised to 250MHz, considerations must be made for the Hypertransport speed to fall in the range of 1.8GHz. The memory timings should also be more relaxed and select a ratio e.g DDR2-533 to ensure that memory is not the bottleneck to overclocking.

In the following page, we present some of the screen shots of running the CPU+board at 3.8GHz.

Settings used

{Advanced -> CPU Configuration}
Multiplier/Voltage Change => Manual
CPU Frequency Multiplier => x19.0 3800 MHz
CPU Voltage => 1.5250 V
NB Frequency Multiplier => x9.0 1800 MHz
NB Voltage => 1.3500 V

{Advanced -> Chipset Configuration}
Advanced Clock Calibration => Auto
mGPU => 1.40 V

edit 2: now look at the CPU Vantage score in OCed mode,the scaling is very bad.Wth is going on with their scores,how can they miss this(they can't be such n00bs,can they? j/k). The CPU score @3.8Ghz is 11471pts while at default 3Ghz is ~10050pts ,that's a massively low score for OCed speed of 3.8GHz.The 11471pts is what a Phenom II @ 3.44Ghz would get ,not at 3.8Ghz.

[G-UNIT91]

Hey people, overclockworkbench seems to just broke the NDA(if they ever signed one that is),and tested Phenom II 940 vs QX9770 vs Core i7 940:
http://my.ocworkbench.com/2008/asroc..._X4-info-1.htm

Site is opening slow atm,but Phenom II results seem pretty good,considering the fact both intel QC chips ran at higher clock(QX9770 is 3.2Ghz CPU while Core i7 mostly turbos up to 3066Mhz)

edit: Cinebench 10 32bit scores seem to be way low for their Phenom II system,900-1000pts lower.In 32bit Phenom II @3Ghz should score ~10500 and not ~9600pts... I can't say what's going on there,but their results are hindered in some way in this test(only?),

if it's true......

[malice85]

You're right some of these are kinda strange, but nevertheless if these gaming results turn out to be true then that's more than I expected tbh. Can't wait to get one

[Enoc]

tests

result looks good...but i think they could have done better...memory timmings wise...it's not like they are going to gain...but they could of run 5-5-5-15/12 1066...5-7-7-24

i think DDR3 is going to make the difference in phenom 2...if they really gain that 5% from the transition of DDR2...

[informal]

Guys look at my edits in the first post and you will see something very strange is going on with their Phenom II system.The results are all over the place,OCing scaling is simply bad.I say they have a poor BIOS support or something is playing a trick on them(maybe OS or a driver issue).

[SoF]

oh dear what is going on with all the sites? what are they thinking? "if we can't show best results after the top benchers have their hands on, we simply break NDA to be first and got our hits anyway" ???

[informal]

Yeah SoF,they just did exactly what you said... Anyway,their results are pretty inconsistent and look like they were done in a hurry(your comment rings true even more ). Look at the CPU vantage scaling with the OC,simply bad-either a poor BIOS support,issue with an OS or simple bad "review" .They could just wait like everyone else and published the proper scores at the agreed time...

[chew*]

Well at least despite somewhat take it with a grain of salt results via this test what we saw in the crysis comparison that amd had on hand seems to ring true.

I'll wait for more reviews to get released.

[Zucker2k]

OCW:

F.E.A.R is set to High/High(1600x1200) for the test and it scored 90 fps. On both Core i7 and QX9770 platform, both also scored ~90-91 fps. It looks like this game is GPU bound, thus no matter how fast the CPU is, the GPU isn't able to go any faster.

How do you test a cpu and make the test gpu bound?

AMD Phenom II X4 scored 51 for Snow benchmark and 70 for Cave Benchmark. On the QX9770 running on P45XE-WiFi-N, it only scored Snow 49 and Cave 67. Looks like the Phenom II X4 940 is a winner in this test.

Looks like they're not sure of their own results as almost every commentary is preceded by this remark.

Sorry OP, by your own admission, something fishy is going on here. I mean you don't need special glasses to see through this one. My point being, this thread is nothing but a propagation of propaganda since you yourself has admitted there are many things wrong with those results; I'll be bolder and say, all the tests are questionable on the basis of their testing criteria alone. FUD couldn't do better to generate hits.

[Oliverda]

The 3DM06 CPU score is pretty low. I get more with my 9950@3GHz. I think that something was wrong with the testing method or the setup.

[informal]

My point being, this thread is nothing but a propagation of propaganda

Yeah,couldn't expect less from you

The Lost planet test(one of the two) is CPU bound,you can check this in intel section if you want.
The test is fishy not because of the reason you said(fear test is the poorest reason i can think of).I detailed in my edits what's fishy.The OC scores don't scale with MHz,cinebench is 1000pts lower than it should be.All this tells us they messed up somewhere and broke the NDA for nothing.

[chew*]

Well the first mistake was not loading the cpu's up enough to eliminate the gpu bound issues. They should have at least used 1 4870x2. That would be the start of creating a bottleneck. Running 2 would definitely create a bottleneck somewhere, then we might see a wider gap between one platform or the other.

Is there a page that i'm missing that shows what all 3 configs are using save the cpu and ram?

[AbelJemka]

OCW:
Sorry OP, by your own admission, something fishy is going on here. I mean you don't need special glasses to see through this one. My point being, this thread is nothing but a propagation of propaganda since you yourself has admitted there are many things wrong with those results; I'll be bolder and say, all the tests are questionable on the basis of their testing criteria alone. FUD couldn't do better to generate hits.

Come on, you blame Informal because OCW test sucks?

[FlanK3r]

I hope better OC on AIR at ASUS M3A-T deluxe, right?
I seen here any screens cinebench 64bit and it was more better

[Zucker2k]

Come on, you blame Informal because OCW test sucks?

I don't blame fim for anything, I'm just pointing out that this sucky and very questionable "review" does not need a thread since the results are less than legit. A thread was opened yesterday about this and was closed after it was established that the results were far off: http://www.xtremesystems.org/forums/...d.php?t=211253

Besides this, Informal had already brought up this review in another thread, so why create a new thread and then proceed to condemn the results? It makes no sense? Does it to you?

[informal]

Besides this, Informal had already brought up this review in another thread, so why create a new thread and then proceed to condemn the results? It makes no sense? Does it to you?

Because that post in the review thread was lost in a sea of spam posts which you gladly contribute lately?
I condemned the results because the "review" was full of errors and had inconsistent numbers all over the place?Does it makes sense now?Ah yes it does.
You can shrug all day long,but to me you contribute zero to this section(and i can bet a majority agrees with me on this).

[Zucker2k]

Because that post in the review thread was lost in a sea of spam posts which you gladly contribute lately?
I condemned the results because the "review" was full of errors and had inconsistent numbers all over the place?Does it makes sense now?Ah yes it does.
You can shrug all day long,but to me you contribute zero to this section(and i can bet a majority agrees with me on this).

Thank you for my making my point for me; you wouldn't have taken that coming from me. So, why create a new thread for what obviously is to quote you, a review that was:

full of errors and had inconsistent numbers all over the place?

[Enoc]

Is there a page that i'm missing that shows what all 3 configs are using save the cpu and ram?

no, there is no such page...but the when you look at the whole article...the phenom 2 config is:

MB ASRock AOD790GX/128M
2X1GB DDR2-1066 at C5-7-7-24-2T
ATi Radeon HD 4850

and they point out that the core i7 was running 3gb of DDR3.

[Halk]

Why'd they run it on GX rather than FX? Makes no sense!

[informal]

Thank you for my making my point for me; you wouldn't have taken that coming from me. So, why create a new thread for what obviously is to quote you, a review that was:

full of errors and had inconsistent numbers all over the place?

Lol you can't get it.As you can see i didn't see the errors until i edited my posts.I first posted it in the review thread and it was at the bottom of the page(barely noticeable in the sea of other posts).After that i posted it here where it is clearly visible.But as you saw in my first post,after more reading(since the site was slow as a snail) i saw the fishy numbers and poor OC scaling and hence i edited my post 2 times.

Gee i wish i was a psychic and i could see all the flaws of the review instantly before i can get to actually read it thru. But your mission of trolling is never ending so why should you stop now,it's near the finish line(launch day) and you need to shift in higher gear

[chew*]

Why'd they run it on GX rather than FX? Makes no sense!

That really doesn't matter. performance is negligible imo 790 fx versus gx. More importantly they used a asrock. I'd rather see a tier 1 manufacturer.

On a side note guys

[AbelJemka]

I don't blame fim for anything, I'm just pointing out that this sucky and very questionable "review" does not need a thread since the results are less than legit. A thread was opened yesterday about this and was closed after it was established that the results were far off: http://www.xtremesystems.org/forums/...d.php?t=211253

Besides this, Informal had already brought up this review in another thread, so why create a new thread and then proceed to condemn the results? It makes no sense? Does it to you?

I understand but it's mod job to judge if a thread is worthless not yours.
That i don't understand it's why you spend so much time here posting negatives comments on a product you will obviously never buy

[Zucker2k]

Lol you can't get it.As you can see i didn't see the errors until i edited my posts.I first posted it in the review thread and it was at the bottom of the page(barely noticeable in the sea of other posts).After that i posted it here where it is clearly visible.But as you saw in my first post,after more reading(since the site was slow as a snail) i saw the fishy numbers and poor OC scaling and hence i edited my post 2 times.

Gee i wish i was a psychic and i could see all the flaws of the review instantly before i can get to actually read it thru. But your mission of trolling is never ending so why should you stop now,it's near the finish line(launch day) and you need to shift in higher gear

Don't worry about me trolling; I'm not a troll, I just call things as they are, I know that's hard for you to see past. I'm not here to praise blindly. I'm out of this thread which is "full of errors and had inconsistent numbers all over the place."

Lesson: Please read the review before you create a thread about it; I mean you're not getting paid to do this, or are you? OUT!

[Titan7171]

why are their voltages set to 1.368v for 3ghz?? thats bs I can get 3ghz@ 1.298v on Phenom now

[kemo]

Why the rush ,, i don't mind waiting for NDA to be lifted