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

[Darth Farter]

Read:

http://www.theinquirer.net/?article=41970


if true this could indeed explain the Dead silence around it.... 3ghz Agena FX with 2 hd2900xt's breaking 30.000 marks in 3dmark2006

how much does a core2 fare atm?

 

[JAG87]

I think I have some alaskan ice melter left in the garage. Let me get a shovel so you can feed it to me while I read the latest inquirer article.



edit

this is how you know inquirer numbers are BS

http://www.theinquirer.net/?article=41970

Windows Me II (Vista) Ultimate was installed as the operating system of choice. For some odd reason, the 32-bit version was installed in a system with 4GB of memory, needless to say the system detected 3.24GB, and benchmarketing commenced.

http://support.microsoft.com/kb/929605

The reduction in available system memory depends on the devices that are installed in the computer. However, to avoid potential driver compatibility issues, the 32-bit versions of Windows Vista limit the total available memory to 3.12 GB. See the "More information" section for information about potential driver compatibility issues.

plus everyone knows that with two 512MB cards you will have less than 3GB of memory available in win32. I think Theo Valich needs a new prescription, cause hes reading one number for another.

 

[swtethan]

well its possible

if they were running 640x320 resolution! (which they probably did) they did not disclose anything about the benchmark settings.

 

[ultra laser]

I have faith in AMD to come out with a real screamer. They've never disappointed in the past. Besides, their native quadcore solution is just so much more elegant than intel's. I'd be really surprised if it can outperform it.

 

[nyker96]

somehow I don't believe this guy. partly since AMD is in such finacial trouble that even a little leak of good results could boost investor confidence and draw in much needed cash. This is probably not true.

 

[bryanW1995]

isn't 3dmark 06 more dependent on the gpu settings than cpu? the current record is just over 27000, a 3000 pt increase in cpu alone would be, uh, incredible...

 

[Idontcare]

Originally posted by: JAG87
this is how you know inquirer numbers are BS

http://www.theinquirer.net/?article=41970

Windows Me II (Vista) Ultimate was installed as the operating system of choice. For some odd reason, the 32-bit version was installed in a system with 4GB of memory, needless to say the system detected 3.24GB, and benchmarketing commenced.

http://support.microsoft.com/kb/929605

The reduction in available system memory depends on the devices that are installed in the computer. However, to avoid potential driver compatibility issues, the 32-bit versions of Windows Vista limit the total available memory to 3.12 GB. See the "More information" section for information about potential driver compatibility issues.

plus everyone knows that with two 512MB cards you will have less than 3GB of memory available in win32. I think Theo Valich needs a new prescription, cause hes reading one number for another.

Those were my immediate thoughts as well JAG87...somebody got their stories mixed up at the good old Inquirer.

 

[HopJokey]

Originally posted by: ultra laser
I have faith in AMD to come out with a real screamer. They've never disappointed in the past. Besides, their native quadcore solution is just so much more elegant than intel's. I'd be really surprised if it can outperform it.

What does having a monolithic die for all four cores have to do with single threaded performance?

 

[Sledgehamer70]

The story holds merit, and will be shot down or up once more benchmarks are revealed. Theo hasn?t been wrong about much if anything in the past and hasn?t had a track record like ole Charlie at the INQ. Also AMD quad core will benefit GPU's more as everything is on die and no bus to slow it down. So the increase could mean faster timings within the CPU and GPU, but not mean the CPU is faster than Intel in other app's.

 

[Lord Banshee]

Not favoring either side, but, 3dmark06 has a single threaded and multi threaded cpu bench in its suite that runs by default and i am sure if these scores are right it is with the default benchmark settings.

 

[Keysplayr]

Originally posted by: HopJokey

Originally posted by: ultra laser
I have faith in AMD to come out with a real screamer. They've never disappointed in the past. Besides, their native quadcore solution is just so much more elegant than intel's. I'd be really surprised if it can outperform it.

What does having a monolithic die for all four cores have to do with single threaded performance?

Nothing. But Dual/Quad cores do score higher in 3DMark06 over a single core. So apparently, more cores means higher scores. Whether or not the cores are native is pretty much irrelevant as you can see.

 

[SickBeast]

Hopefully 3ghz won't be out of reach for the retail chips...

If the thing comes out at 2ghz like everyone is saying, then really it scores 20,000 which is underwhelming.

Interesting results nonetheless...I look forward to a 'proper' review.

 

[JAG87]

For general information


From the Inquirer:

To be clear with you, dear readers, two HD2900XT cards were overclocked to 830MHz for the core and 900MHz for the memory, but ominous Catalyst 7.7 drivers were used.

From NordichHardware:

The GPUs were overclocked to 880MHz and the memories to 950MHz, with modified memory latencies for better performance.

Take the scores from NordichHardware (which are higher because the clocks are higher) and running the 3dMark06 score calculator shows that even a CPU score of 99999 (which is the maximum the calculator allows) does not yeild a score of 30K. The calculator has been tested, and it always reports the correct overall scores. This goes to show that the formula used to calculate overall score is largely based on the graphic scores not the CPU score.

So Mr. Theo from the Inquirer, unless AMD was running next gen video cards, can you please explain to us noobs how you get a score of 30K? Are you gonna tell us that a 3.73 GHz kentsfield is bottlenecking those cards?

 

[swtethan]

Originally posted by: JAG87
For general information


From the Inquirer:

To be clear with you, dear readers, two HD2900XT cards were overclocked to 830MHz for the core and 900MHz for the memory, but ominous Catalyst 7.7 drivers were used.

From NordichHardware:

The GPUs were overclocked to 880MHz and the memories to 950MHz, with modified memory latencies for better performance.

Take the scores from NordichHardware (which are higher because the clocks are higher) and running the 3dMark06 score calculator shows that even a CPU score of 99999 (which is the maximum the calculator allows) does not yeild a score of 30K. The calculator has been tested, and it always reports the correct overall scores. This goes to show that the formula used to calculate overall score is largely based on the graphic scores not the CPU score.

So Mr. Theo from the Inquirer, unless AMD was running next gen video cards, can you please explain to us noobs how you get a score of 30K? Are you gonna tell us that a 3.73 GHz kentsfield is bottlenecking those cards?

hey youre right, even with 8k sm2, 8k hdr and 90k cpu, you only get 23k points.

taking the scores from this run http://service.futuremark.com/compare?3dm06=2526811

you need a 28,000 point cpu vs the 5.0GHz quad score of 7400 to get 30,000 marks



so this means........

1. amd has a chip 3x faster than an intel quad @ 5GHz

2. article is lieing

3...?

 

[MarcVenice]

Erm, I think windows is still seeing 3.12 gb and not less then 4gb because even though there are 2 cards in crossfire, only 512mb total gets adressed each frame. Not the full 1024mb. I might be wrong though, since I can't find real information on how Crossfire works. I know thats somewhat how SLI works though. I wouldn't dismiss the information solely based on how much memory is being adressed.

 

[JAG87]

Originally posted by: swtethan

Originally posted by: JAG87
For general information


From the Inquirer:

To be clear with you, dear readers, two HD2900XT cards were overclocked to 830MHz for the core and 900MHz for the memory, but ominous Catalyst 7.7 drivers were used.

From NordichHardware:

The GPUs were overclocked to 880MHz and the memories to 950MHz, with modified memory latencies for better performance.

Take the scores from NordichHardware (which are higher because the clocks are higher) and running the 3dMark06 score calculator shows that even a CPU score of 99999 (which is the maximum the calculator allows) does not yeild a score of 30K. The calculator has been tested, and it always reports the correct overall scores. This goes to show that the formula used to calculate overall score is largely based on the graphic scores not the CPU score.

So Mr. Theo from the Inquirer, unless AMD was running next gen video cards, can you please explain to us noobs how you get a score of 30K? Are you gonna tell us that a 3.73 GHz kentsfield is bottlenecking those cards?

hey youre right, even with 8k sm2, 8k hdr and 90k cpu, you only get 23k points.

taking the scores from this run http://service.futuremark.com/compare?3dm06=2526811

you need a 28,000 point cpu vs the 5.0GHz quad score of 7400 to get 30,000 marks



so this means........

1. amd has a chip 3x faster than an intel quad @ 5GHz

2. article is lieing

3...?

Card clocks 1080/950

you did realize that right swtethan.

Originally posted by: MarcVenice
Erm, I think windows is still seeing 3.12 gb and not less then 4gb because even though there are 2 cards in crossfire, only 512mb total gets adressed each frame. Not the full 1024mb. I might be wrong though, since I can't find real information on how Crossfire works. I know thats somewhat how SLI works though. I wouldn't dismiss the information solely based on how much memory is being adressed.

you are wrong marc, because what matters is the total adressable memory on the pci bus. not what crossfire or SLI can use (in which case you are right). the memory seen in windows would most likely be around 2.75 GB

 

[ayabe]

Pretty amazing if accurate. I'm pretty sure theINQ knew this would create some controversy and would be sure to make sure they knew what they were talking about.

Oh well, guess we'll find out for sure in a few weeks.

 

[Accord99]

Originally posted by: Sledgehamer70
Theo hasn?t been wrong about much if anything in the past and hasn?t had a track record like ole Charlie at the INQ.

Theo's the one who started the infamous reverse-HyperThreading rumors:

http://www.theinquirer.net/?article=32589

 

[Dainas]

Well, atleast will make for one heck of a laugh if there was drink before the viewing and these scores turned out to be 3Dmark05. Sure why not though, its not as if AMD has not put a mass of effort into this new architecture... I would hate for fanboys to have their blind faith validated But AMD deserves to be back on top.

 

[dmens]

Originally posted by: Accord99

Originally posted by: Sledgehamer70
Theo hasn?t been wrong about much if anything in the past and hasn?t had a track record like ole Charlie at the INQ.

Theo's the one who started the infamous reverse-HyperThreading rumors:

http://www.theinquirer.net/?article=32589

LOL, that seals the deal. Guy is a complete quack.

 

[NoSoup4You]

I would love for that article to be true, if only because AMD has been important in keeping Intel honest... but the manner in which the author wrote that article leads me to believe it's totally bogus. Combined with that 'Reverse Hyper Threading' post and I'll just have to see it to believe it.

 

[bryanW1995]

Originally posted by: ayabe
Pretty amazing if accurate. I'm pretty sure theINQ knew this would create some controversy and would be sure to make sure they knew what they were talking about.

Oh well, guess we'll find out for sure in a few weeks.

are you serious? does the inq ever know what they're talking about?
Their web page title says " news, reviews, facts, and fiction". I wonder which category this one falls into...

 

[imported_jX]

Originally posted by: bryanW1995
Their web page title says " news, reviews, facts, and fiction". I wonder which category this one falls into...

I'm fairly sure it says "news, reviews, facts and friction".

 

[bryanW1995]

lol

 

[Laminator]

Here's a good tip: never read The Inquirer. They're reliably unreliable. I think most people have pretty much learned to ignore everything they say by now.