A comparison article is:
AMD's 3GHz K10 to break 30,000 3DMark06
(Just keeping to the facts, and nothing but the facts please. ;-) )
Regards,
Martin

LOL - you are linking to a GFX-card bechmark without any documentation (screenshots)!
Do we have any use for this here at S&H?

Stick to the cold facts - plz!

Kiva

Hmm

I thought Cinebench was a rendering program using only CPU or am i wrong there, the rendered picture is only displayed in a regular 2d fashion what the cpus was rendering.

That was what i thought Cinebench really were.

Ofc it can use 3d modelling too but when rendering final stages the cpu should only be involved.

Correct me if im wrong..

//Vyper

---

_________________________________________________________________________
Addicted to SETI crunching!
Founder of GPU Users Group

Understood Who?

A comparison article is:
AMD's 3GHz K10 to break 30,000 3DMark06
(Just keeping to the facts, and nothing but the facts please. ;-) )
Regards,
Martin

LOL - you are linking to a GFX-card bechmark without any documentation (screenshots)!
Do we have any use for this here at S&H?

Stick to the cold facts - plz!

Kiva

---

Greetings from Norway

Crunch3er & AK-V8 Inside

LOL . pooo pooo pi doooo
I go to bed, I finish my perf tuning for the day ... 4am, 8am in office :(
outch!

who?

Heh

As i said in the 30K 3dm06 post i mentioned we will see in real world app how it would perform and so.
Competition is good for both it drives technologies forward instead of riding the wave..

From this view it surely seems like that K10 is not a good Cinebench performer from what i can see, i don't expect K10 to be 100% victorious but i think its faster in certain conditions..
It's all up to comparing apples & pears atm, they're rather alike but very different indeed.

Only time will tell , but we seem now to know for sure that Cinebench is not K10's strong part atm..
Very interested on Penryn perfomance atm and SSE4 optimised apps -> s@h client.

Yum yum.. :)

Perhaps i would PM you so u could send me one for free ;-) to bench the crap out of it in all aspects heat/noise/power ratio .. I'm in the PC OEM business and in my twelve years employment i've never ever received anything from any company so they would endorse their new technologies to sell to customers..

Kind Regards Vyper

well, on the rest of the post, you can get SuperPI and Wprime, fritz chess...
some look ok too.

who?

Who?, I know you probably don't know or are not Allowed to answer this due to a NDA.

But i'll ask you anyway. As your clearly in the know on many things, Intel. Do you have any idea when the Mac Pro will be updated?

If I answer this, i ll be cut in little pieces, blend, liquify, then, i ll be fired :)

who?

Who?, I know you probably don't know or are not Allowed to answer this due to a NDA.

But i'll ask you anyway. As your clearly in the know on many things, Intel. Do you have any idea when the Mac Pro will be updated?

LOL . pooo pooo pi doooo
I go to bed, I finish my perf tuning for the day ... 4am, 8am in office :(
outch!

who?

Heh

As i said in the 30K 3dm06 post i mentioned we will see in real world app how it would perform and so.
Competition is good for both it drives technologies forward instead of riding the wave..

From this view it surely seems like that K10 is not a good Cinebench performer from what i can see, i don't expect K10 to be 100% victorious but i think its faster in certain conditions..
It's all up to comparing apples & pears atm, they're rather alike but very different indeed.

Only time will tell , but we seem now to know for sure that Cinebench is not K10's strong part atm..
Very interested on Penryn perfomance atm and SSE4 optimised apps -> s@h client.

Yum yum.. :)

Perhaps i would PM you so u could send me one for free ;-) to bench the crap out of it in all aspects heat/noise/power ratio .. I'm in the PC OEM business and in my twelve years employment i've never ever received anything from any company so they would endorse their new technologies to sell to customers..

Kind Regards Vyper

---

_________________________________________________________________________
Addicted to SETI crunching!
Founder of GPU Users Group

It's certainly not going to get much in the way of credit, if it's got 64b linux, and an official 64b client on it using 5.12 instead of 5.27, or hey, here's an idea...an actually optimized 64b linux app. That's for sure. lol

"the above humourous post presupposes that the 32b 5.27 linux app is optimized in some fashion to a higher degree than it's 32b 5.12 counterpart. ATM, I have no supporting evidence of that." LOL

LOL . pooo pooo pi doooo
I go to bed, I finish my perf tuning for the day ... 4am, 8am in office :(
outch!

who?

PS: I don t think you know who you just called "naive" ... Every overclocker on the planete respect Ed!

Then fully explain the numbers for both articles...?

Does Cinebench use full optimisations for both Intel and AMD?

Regards,
Martin

Do you really think AMD will be demonstrating Cinebench if it was putting them into a dissavantage? Here, they use it: you tube video of CINEBENCH running ON BARCELONA FROM AMD MARKETING!!!!

I don't think they are stupid enough to use something putting them into dissavantage. do you?

Cinebench is a fair test, live with it!
I have big respect for them, please do not think they are that bad.

who?

PS: I don t think you know who you just called "naive" ... Every overclocker on the planete respect Ed!

Then fully explain the numbers for both articles...?

Does Cinebench use full optimisations for both Intel and AMD?

Regards,
Martin

---

See new freedom: Mageia Linux
Take a look for yourself: Linux Format
The Future is what We all make IT (GPLv3)

It's certainly not going to get much in the way of credit, if it's got 64b linux, and an official 64b client on it using 5.12 instead of 5.27, or hey, here's an idea...an actually optimized 64b linux app. That's for sure. lol

"the above humourous post presupposes that the 32b 5.27 linux app is optimized in some fashion to a higher degree than it's 32b 5.12 counterpart. ATM, I have no supporting evidence of that." LOL

A comparison article is:

AMD's 3GHz K10 to break 30,000 3DMark06
(Just keeping to the facts, and nothing but the facts please. ;-) )

I was waiting for this one:
here is what one of the best overclocker think about it: http://www.overclockers.com/tips01208/ (yes, Theo is kind of strange on this one, the mathematics are not working on this...)

That article is 'grossly naive' and misleading. No mention is made on whether the tests are CPU limited, system RAM limited, or graphics cards limited. ("limited" meaning utilization 'maxed out'.)

Note also that the Inquirer tests were done with overclocking a dual graphics card system... AMD also enjoys higher system RAM bandwidth than the Intel FSB -> Northbridge -> RAM path.

Cinebench is a public test, use to be one of the favorit AMD test on K8, same for superPI.

That's fine. Have they used the Intel compiler that 'just happens to' switch OFF the SIMD optimizations if the CPU is found to be non-Intel?

again, i just provide the pointers, i did not run anything myself.

pooo pooo pi doooo.

A more useful performance test is to compare like-for-like optimizations on comparable systems. All we have at the moment is similarly ghostly smoke and leaks from both AMD and Intel of 'future promises' sometimes...

All good for intrigue and conjecture and dreaming.


And I didn't know you were a Betty Boop (sp?) fan!

Regards,
Martin

http://www.futuremark.com/community/halloffame/
As long as i don t see the 3dMark score there, it is science fiction.
if somebody was able to get 30K, trust me, he will file it!


waiting for this, we have the nice performance of Cinebench on 64bits, where AMD is suppose to be at its best ... pooo poooo pi dooo.
As you probably know, on 64bits, it is not x87, it is SSE2 and they used it for public demo recently, so, i don t think somebody can say it is an unfair test.
If you need a point of comparaison, here it is: http://www.legitreviews.com/article/521/1/ (ran with 20 journalists checking the machine and looking)

who?
PS: I don t think you know who you just called "naive" ... Every overclocker on the planete respect Ed!

A comparison article is:

AMD's 3GHz K10 to break 30,000 3DMark06
(Just keeping to the facts, and nothing but the facts please. ;-) )

I was waiting for this one:
here is what one of the best overclocker think about it: http://www.overclockers.com/tips01208/ (yes, Theo is kind of strange on this one, the mathematics are not working on this...)

That article is 'grossly naive' and misleading. No mention is made on whether the tests are CPU limited, system RAM limited, or graphics cards limited. ("limited" meaning utilization 'maxed out'.)

Note also that the Inquirer tests were done with overclocking a dual graphics card system... AMD also enjoys higher system RAM bandwidth than the Intel FSB -> Northbridge -> RAM path.

Cinebench is a public test, use to be one of the favorit AMD test on K8, same for superPI.

That's fine. Have they used the Intel compiler that 'just happens to' switch OFF the SIMD optimizations if the CPU is found to be non-Intel?

again, i just provide the pointers, i did not run anything myself.

pooo pooo pi doooo.

A more useful performance test is to compare like-for-like optimizations on comparable systems. All we have at the moment is similarly ghostly smoke and leaks from both AMD and Intel of 'future promises' sometimes...

All good for intrigue and conjecture and dreaming.


And I didn't know you were a Betty Boop (sp?) fan!

Regards,
Martin

---

See new freedom: Mageia Linux
Take a look for yourself: Linux Format
The Future is what We all make IT (GPLv3)

You can finally make your own opinion: here
what matter are the screen shot.

What are we comparing with?

And the first screenshot shows a test for 32 bit operation?...

And are those tests with code optimised for the Intel CPUs?... (And so the SIMD instructions are not utilised?)

It looks good that AMD have rushed through their next revision so quickly.


A comparison article is:

AMD's 3GHz K10 to break 30,000 3DMark06



(Just keeping to the facts, and nothing but the facts please. ;-) )

Regards,
Martin

I was waiting for this one:
here is what one of the best overclocker think about it: http://www.overclockers.com/tips01208/ (yes, Theo is kind of strange on this one, the mathematics are not working on this...)

Cinebench is a public test, use to be one of the favorit AMD test on K8, same for superPI.

again, i just provide the pointers, i did not run anything myself.

pooo pooo pi doooo.

who?
EDIT: I forgot, I believe screen shots, i don t believe when somebody tell me that 2 weeks ago, i saw a world records, but, i can t remember exactly where and what was it ....

You can finally make your own opinion: here
what matter are the screen shot.

What are we comparing with?

And the first screenshot shows a test for 32 bit operation?...

And are those tests with code optimised for the Intel CPUs?... (And so the SIMD instructions are not utilised?)

It looks good that AMD have rushed through their next revision so quickly.


A comparison article is:

AMD's 3GHz K10 to break 30,000 3DMark06

(Just keeping to the facts, and nothing but the facts please. ;-) )

Regards,
Martin

---

See new freedom: Mageia Linux
Take a look for yourself: Linux Format
The Future is what We all make IT (GPLv3)

You can finally make your own opinion: here
what matter are the screen shot.
Don't shoot at me, i am only the messenger ;-)

who?