AMD Zambezi news, info, fans !

[duron]

what benchmark software's will guys use when BD finally hit retail?

[FlanK3r]

CInebenchs ofd course, POV-Ray, y-cruncher with AVX support, x264 HD benchmark (4.0), wprime 1.15, Zoner photo studio etc

[zalbard]

Don't forget 3Dmarks, PCMarks, Chess Benchmark, Sandra, maybe AIDA64 benches.

[Madmanden]

Cinebench scores are all I need to decide. :-)

[duron]

thanks, i'll get those ready too

[FlanK3r]

Here is is complete:
DosBOX
superpi 1M
Hyperpi 32M
Maxxpi 4M
MaxxMem
MaxxFlops
AIDA Queen
AIDA Hash
AIDA Memory test
Geekbench
Fritchess bench
wprime 32M and 1024M
y-cruncher (set 0,1,3=536MB)
winrar bench
7-zip (comprimation and decomprimation)
Zoner Photo Studio, coverting photos (na 1024x768, supersampling,*.jpg)
Cinebench R10 (single + multi)
Cinebench R11.5
POv-Ray 3.7 (all CPUs)

x264 3.0
x264 4.0
AVS video konverze (AVP game trailer video to HD video 720p:MPEG4 4500kb/s audio mp3 192kbps)
ConvertXtoDVD (avi to DVD mvoie Inside job)
mediaespresso (convert-mp4.DivX 640x480)


3D Vantage CPU

megatasking:
Cinebench R11+Maxxpi 64M+winrar test, results from R11.5 and Maxxpi only!


PS: I could add next some practice sowt experience or benchmarks, sugest me

[RaV[666]]

OBR updated his blog with some info.He claims 8150P is going to cost 269$, that would mean targeting 2500K as competitor :-/ .
In separate news he confirms 6core 2011 sandy is going to go for 560$ which neatly puts it out of my price range ... :-( ,if BDv1 wont deliver i might wait for komodo ...

[TheBreezyBB]

But, the question is who believes OBR when it comes to AMD news?

[Assimilator87]

what benchmark software's will guys use when BD finally hit retail?

I'm waiting to see some F@H numbers cuz I'm hoping Bulldozer will be an absolute beast at bigadv due to the dedicated integer cores instead of Hyperthreading.

[RaV[666]]

But, the question is who believes OBR when it comes to AMD news?

Well, when he leaked, he published better numbers than these chinese guys at first plus gave out a logical explanation for their scores.He obviously has access to ES`s also, both intel and amd.
He mostly screws with people by for example "publishing" scores which are aggregate of many individual ones so you still dont know ;-).
He was right about the delay at first, then he was right about october availability (launch on 19th sep means around october world wide availability).
There are other instances.I concur that the guy is an ass.But hes only source on BD now.

[Dresdenboy]

Yesterday I tweeted a link to one of OBR's CPU reviews for PC Tuning CZ, where he explains his way of calculating this mixed score and also lists the used apps again:
http://translate.google.com/translat...5%3Fstart%3D11

In that review he set each max score of a benchmark to 100 and calculated ratios for all lower scores in this benchmark. Finally he simply creates an average value for those scores. In the leaked results he seems to have scaled the results again so that the highest score becomes 100. While one problem with this method might be the arithmetic mean being used (instead of a geometric mean as for SPEC) and the missing weights since SuperPi 1M or Quake running in the DOSBox emulator (very likely single threaded) becomes as important for the score as more likely everyday tools like TrueCrypt, 7zip or Excel.

@RaV[666]:
I'd even assume that not everything OBR called "fake" afterwards was a fake. As things developed, some posters at p3dnow and me already assumed that some NDA trouble is going on behind the scenes, which became apparent after he put some explainations on his blog and removed pics. So his "fake call" of DH slides might simply be a reaction to that, because there is a lot of plausible stuff in these "faked" pics, e.g. you couldn't easily exchange numbers in an AIDA64 mem bench screen with all those background colors being untouched. Further it would be too much work just to create this bunch of screens to have some fun. For some part I was involved in the AMDZone April 1st article, so I know what I'm talking about.

[SEA]

I can't believe OBR's results taken ANY seriously...
They look highly suspicious and absolutely non-adequate, even IF he really has some chip... And even it is bulldozer maybe... Not easy to believe but still maybe, say, he got a BD with turned off blocks, buffers and caches and heavily treated with microcode to bring chip to life (and taped somewhere in winter)...
Is there any leaked perf. numbers except those from OBR at all?

[Dresdenboy]

I can't believe OBR's results taken ANY seriously...
They look highly suspicious and absolutely non-adequate, even IF he really has some chip... And even it is bulldozer maybe... Not easy to believe but still maybe, say, he got a BD with turned off blocks, buffers and caches and heavily treated with microcode to bring chip to life (and taped somewhere in winter)...
Is there any leaked perf. numbers except those from OBR at all?

I'm sure most people here know about the state of AMD BD ES. There were benchmark leaks in some Chinese forums already. And those also contained AIDA scores, which revealed cache problems. So if you keep all that in mind it's not a problem to check if there is some truth in OBR's BD ES fumblings.

[RaV[666]]

I can't believe OBR's results taken ANY seriously...
They look highly suspicious and absolutely non-adequate, even IF he really has some chip... And even it is bulldozer maybe... Not easy to believe but still maybe, say, he got a BD with turned off blocks, buffers and caches and heavily treated with microcode to bring chip to life (and taped somewhere in winter)...
Is there any leaked perf. numbers except those from OBR at all?

There has been two or three leaks (beside donanhimber and OBR) with B0 ES chips which showed WAY worse performance figures, OBR stated that their scores were with non fully functional bioses as he got similar scores on other boards than crosshair V with newest bios.
OBR apparently had B1 silicon which clocked to 4.8ghz stably and was apparently without any bugs.As of now he claims to have B2 retail silicon, but hes not made any benches or OC sessions with it.
You gotta take into perspective BDv1`s bug ,which will be fixed with komodo ,its officialy stated by AMD in its papers, in some circumstances it has way worse cache performance (as i understood it at least). AES performance also is not going to be as good as sandys, as amd stated it also in some slide.So there are definitely some drawbacks to BD in its current state (up to B1 at least).

[SEA]

OBR just slowly making up the news from mind to keep attention all the time. That's why he continuously shows B0, B1 and now B2. He slowly "releases" news and "benches" by small dozes of info.
Yeah, he got a B2 silicon, but he did not do any tests or overclocks... the 1M SuperPI takes weeks to run

"Banana totale obosratus"...

[undone]

hey, is there ANY necessary to discuss OBR all along? He lost credit ever, no matter either what he said is true or not. What you did can only increase his blog's traffic at all.

Yeah, he got a B2 silicon

Prove?

[SEA]

hey, is there ANY necessary to discuss OBR all along? He lost credit ever, no matter either what he said is true or not. What you did can only increase his blog's traffic at all.


Prove?

Again, my point is exactly same like yours: - OBR's news is total BS. He is slowly fabricating news to draw public attention.
The clownery "he got B2 silicon" goes from there - he states that he got it

[RaV[666]]

OBR just slowly making up the news from mind to keep attention all the time. That's why he continuously shows B0, B1 and now B2. He slowly "releases" news and "benches" by small dozes of info.
Yeah, he got a B2 silicon, but he did not do any tests or overclocks... the 1M SuperPI takes weeks to run

"Banana totale obosratus"...

He did a superpi runs 3 times already (of course if you believe him).While never giving out full score, just a first number.The time he got was no better(or marginally) than phenom II.
Although i dont see a point of doing a superpi at all.
And if you take into account hes a type of person that likes attention.Its not weird why he releases these screens benches the way he does.Read up dresden boy above for a good analysis.
Anyhow, hes the only one with having high probablity of having the chip and willing to share some info.

[SEA]

He did a superpi runs 3 times already (of course if you believe him).While never giving out full score, just a first number

Yeah, he will give 2nd number in one week. In 2 weeks he will give 3rd number...
That's why i call it total BS. Guy invents news for absolutely clear reason.

[SkullCracka]

that person is banned from here .... i say anyone posting links to that place gets banned as well.

stop the trolling....

[Borden]

Is there any hope of ever seeing a dual-FX platform?

[RaV[666]]

Is there any hope of ever seeing a dual-FX platform?

If youre asking about bulldozer one, then according to JF ,No.

[undone]

Anyhow, hes the only one with having high probablity of having the chip and willing to share some info.

NO. AFAIK some sites(like crazy chinese Zol) are much like NDA-Breaker, they are more trustable. If i remember correctly the initial Llano leaking is from them, later they confirmed postponement of zambezi while donanimhaber still said BD release date is Q2.
IMO if they do not leak, no reason to looking forward to other people & sites to do leaking.

[RaV[666]]

NO. AFAIK some sites(like crazy chinese Zol) are much like NDA-Breaker, they are more trustable. If i remember correctly the initial Llano leaking is from them, later they confirmed postponement of zambezi while donanimhaber still said BD release date is Q2.
IMO if they do not leak, no reason to looking forward to other people & sites to do leaking.

They did leak already, they were the first ones.Its just that results were MUCH MUCH worse than OBR`s.
Amount of work OBR would have to put into faking all these screenshots (most were taken down already) ,all those photos, all those benchmarks.
He would have to do pretty much nothing else for 2 months but fake .
And as i presume you think he fakes them in order to bring AMD down, it would make no sense to give out better score leaks than already available, if that would be his premise.
Logically analyzing this i dont think he would put that much work in something that would make a product he presumably hates look better .
Of course that doesnt mean that his results must be in sync with retail products.
Oh ,i would forget, he even sold one, the B0 ES.

[Andi64]

that person is banned from here .... i say anyone posting links to that place gets banned as well.

stop the trolling....

Why? Because you don't like the scores? :S

You can dislike him, he can be banned from here for a reason, but I really really don't think the scores are being modified. He even stated that you should wait for official release and benchmarks, that the scores were from a B1 ES, and he even overclocked his sample to 4.8Ghz on air stating that they clock pretty nicely (and he's an AMD hater they say...)

I don't judge him, I really don't care to. You can like or dislike him. I'm judging the scores and everything he posted on his blog and seems to be very trustworthy.

Anyway, he's selling 3 CPUs:

http://obrovsky.blogspot.com/2011/07...ollectors.html

I would love to buy one, but I'm in Argentina and even a board will be difficult to get. So 200 EUR is too much for me.

I realy want BD to success, I wan't to upgrade my X6 in October, and I really really wish AMD to compete with Intel's ~500us CPUs, that's what I was expecting from AMD after so many years... but best case scenario is BD going against 2600K... and thats boring.

Anyway, I'll wait and hope