really true k10 breaks 30k 3dmark 06 with crossfire?

[SickBeast]

Originally posted by: HopJokey

Originally posted by: spinejam
Ya think it's time to buy AMD stock again? or should i just go to Vegas and put the dough on black!

It would be better to put the money on black at this moment IMHO Right now AMD is losing 600 million dollars a quarter and I wouldn't go near that stock until I see some signs of solid financial stability with the company.

Yeah that's good advice.

Even if Barcelona is a resounding success, it still has to be well-received by consumers and OEMs. Plus it has to have a solid platform. A wait-and-see approach would be wise. You might be able to make a quick short term gain if you buy their stock now, but it's very risky and I would not want it as a long term purchase at this point.

 

[VirtualLarry]

Isn't 3dMark06 more heavy on the GPU than the CPU? I think that such a score is possible, with three (or more?) tweaked-out graphics cards. I think that the CPU doesn't have that much to do with it, although a 3Ghz native quad-core sure would help too.

Edit: You would think that with AMD/ATI being able to break the 30K milestone in '06, that would be something that they would want to scream from the rooftops about, concerning how powerful their CPU+graphics cards are. No matter if they were "tweaked".

 

[SickBeast]

Originally posted by: VirtualLarry
Isn't 3dMark06 more heavy on the GPU than the CPU? I think that such a score is possible, with three (or more?) tweaked-out graphics cards. I think that the CPU doesn't have that much to do with it, although a 3Ghz native quad-core sure would help too.

Edit: You would think that with AMD/ATI being able to break the 30K milestone in '06, that would be something that they would want to scream from the rooftops about, concerning how powerful their CPU+graphics cards are. No matter if they were "tweaked".

They could be hoping intel gets complacent again...

I agree that the benchmark the inquirer used was stupid. The fact that they hyped it so much is even more dumb.

I can't believe noone has run SuperPi on one of these things yet. It could probably do it in under 10 seconds!

 

[Idontcare]

Originally posted by: SickBeast
I can't believe noone has run SuperPi on one of these things yet. It could probably do it in under 10 seconds!

See first post after the OP: http://forums.anandtech.com/me...=2090639&enterthread=y

2GHz K10 sports a 39.6s Spi 1M...not looking good I'm afraid.

 

[Gary Key]

I know what a 2GHz Barcelona does in 3DM06, lets just say that must have been one very very special 3GHz chip and video card setup to hit 30K.

 

[HopJokey]

Originally posted by: Gary Key
I know what a 2GHz Barcelona does in 3DM06, lets just say that must have been one very very special 3GHz chip and video card setup to hit 30K.

LOL, thanks Gary!

 

[Mr Vain]

Originally posted by: Laminator

Originally posted by: Mr Vain
Its amusing seeing die hard Intel fan boys in denial, discrediting individuals and finding all sorts hypothetical reasons etc why AMD K10 cannot run 30K Mark 06 scores.

You fool, no one here is an Intel fan boy. This is the Inquirer we're talking about. It has no positive reputation left.

Only a fool would say that no one here is an Intel fan boy!
I read bullshit in the Inquirer as I read bullshit in the Daily Tech, but sometimes they are right on the money.
You obviously know Jackshit about speculative journalism and how it works.

 

[Arkaign]

Originally posted by: VirtualLarry
Isn't 3dMark06 more heavy on the GPU than the CPU? I think that such a score is possible, with three (or more?) tweaked-out graphics cards. I think that the CPU doesn't have that much to do with it, although a 3Ghz native quad-core sure would help too.

Edit: You would think that with AMD/ATI being able to break the 30K milestone in '06, that would be something that they would want to scream from the rooftops about, concerning how powerful their CPU+graphics cards are. No matter if they were "tweaked".

I think the 'native' multicore thing is overhyped tremendously, and has been ever since the first dual-core X2s and Opterons. Look at dual socket-940 boards, with two single-core Opterons, and they perform more or less identically to a single Opteron dual-core. You might see 2-5% here and there, but the fact remains : as long as both cores are busy under load, it matters not whether the design is 'native' or not.

You're right about the GPU in 3dmark06. The difference between a 8800GTX and a 8800GTS impacts score much more than say, replacing an E6700 with an E6400.

 

[Mitch101]

Its all moot. We will know in 10 days what the dealio is. If you cant wait any longer then do like eric cartman waiting for the Wii and freeze yourself for those 10 days.

I never understood why people get so excited or artificial performance numbers like 3d mark or super pi either.

I would prefer my applications run faster not my synthetic benchmarks and most of these synthetic benchmarks have no real corelation to real world performance. Even if they did achieve 30K in 3d mark it doesnt do any good because there is no game based upon the 3dmark engine.

Game engines paint a better picture of future performance because game companies develop around those engines.

To prove the point doesnt the Opteron memory benchmark outperform the conroe yet real world performance tells the opposite story. They are good at determining which chip designs are superior but real world tells a different story. It just shows that AMD has a better memory interface design it doesnt tell you that Unreal is faster on opteron than conroe. Synthetic benchmarks suck.

 

[mamisano]

Wow, sounds like history repeating itself.... All the fanboys claiming "B.S." on the insane benchmark numbers just prior to the C2D release.

I think we need to keep count of the accuracy of some websites. Lets start keeping count of how many Inquirer stories are true or not.

 

[Arkaign]

Originally posted by: mamisano
Wow, sounds like history repeating itself.... All the fanboys claiming "B.S." on the insane benchmark numbers just prior to the C2D release.

I think we need to keep count of the accuracy of some websites. Lets start keeping count of how many Inquirer stories are true or not.

Well you have a point, but I'm not sure it applies here. I like AMD, have used more AMD processors over the years than Intel ones, so you really can't call me an Intel fanboy by any stretch.

The things that make me think that this is far different from C2D d-day :

(1)- It's 3dmark, cpu speed impacts it very little in comparison to gpu performance, so such a drastic number doesn't make any sense UNLESS the gpu setup is something revolutionary.

(2)- The other leaked benches show merely average cpu performance, at best (The SuperPi), and are much more believable as they are cpu-limited for the most part.

(3)- The resident AMD insiders are playing this very close to the chest, so that leads me to believe that the initial product launch is going to be very weak / and they have little confidence in what is ready to go. We should hope that the product can be ramped up quickly to keep decent competition, as the alternative is unthinkable.

 

[swtethan]

Originally posted by: mamisano
Wow, sounds like history repeating itself.... All the fanboys claiming "B.S." on the insane benchmark numbers just prior to the C2D release.

I think we need to keep count of the accuracy of some websites. Lets start keeping count of how many Inquirer stories are true or not.

it was 20% increase for c2d, not like the 300% increase for barcelona, k?

 

[Mr Vain]

Originally posted by: swtethan

Originally posted by: mamisano
Wow, sounds like history repeating itself.... All the fanboys claiming "B.S." on the insane benchmark numbers just prior to the C2D release.

I think we need to keep count of the accuracy of some websites. Lets start keeping count of how many Inquirer stories are true or not.

it was 20% increase for c2d, not like the 300% increase for barcelona, k?

K10 beating higher Intel Clovertown on floating point by 25%.

http://www.theinquirer.net/default.aspx?article=42090

By Theo Valich: Monday 03 September 2007, 13:23

http://www.vnunet.com/tv/?chan...lipid=1386_vnunet_0096

OUR OWN SISTER SITE, VNUnet.com came up with a video showing AMD Barcelona floating point scores when compared to Intel quad-core Clowertown at sub-90W Thermal envelope. In case of Intel, processor worked at 2.33 GHz (Xeon 5345), while Barcelona worked at 2.0 GHz (Opteron 2350).

 

[Nemesis 1]

Originally posted by: Mr Vain

Originally posted by: swtethan

Originally posted by: mamisano
Wow, sounds like history repeating itself.... All the fanboys claiming "B.S." on the insane benchmark numbers just prior to the C2D release.

I think we need to keep count of the accuracy of some websites. Lets start keeping count of how many Inquirer stories are true or not.

it was 20% increase for c2d, not like the 300% increase for barcelona, k?

K10 beating higher Intel Clovertown on floating point by 25%.

http://www.theinquirer.net/default.aspx?article=42090

By Theo Valich: Monday 03 September 2007, 13:23

http://www.vnunet.com/tv/?chan...lipid=1386_vnunet_0096

OUR OWN SISTER SITE, VNUnet.com came up with a video showing AMD Barcelona floating point scores when compared to Intel quad-core Clowertown at sub-90W Thermal envelope. In case of Intel, processor worked at 2.33 GHz (Xeon 5345), while Barcelona worked at 2.0 GHz (Opteron 2350).

View all the K10 videos you want at Utube thats were they ripped this one from there are many more go watch . Get excited . Not.

 

[Zstream]

Originally posted by: Nemesis 1

Originally posted by: Mr Vain

Originally posted by: swtethan

Originally posted by: mamisano
Wow, sounds like history repeating itself.... All the fanboys claiming "B.S." on the insane benchmark numbers just prior to the C2D release.

I think we need to keep count of the accuracy of some websites. Lets start keeping count of how many Inquirer stories are true or not.

it was 20% increase for c2d, not like the 300% increase for barcelona, k?

K10 beating higher Intel Clovertown on floating point by 25%.

http://www.theinquirer.net/default.aspx?article=42090

By Theo Valich: Monday 03 September 2007, 13:23

http://www.vnunet.com/tv/?chan...lipid=1386_vnunet_0096

OUR OWN SISTER SITE, VNUnet.com came up with a video showing AMD Barcelona floating point scores when compared to Intel quad-core Clowertown at sub-90W Thermal envelope. In case of Intel, processor worked at 2.33 GHz (Xeon 5345), while Barcelona worked at 2.0 GHz (Opteron 2350).

View all the K10 videos you want at Utube thats were they ripped this one from there are many more go watch . Get excited . Not.

I must be nice to you, bah!