ATI Radeon HD 4000 Series discussion

[AliG]

yeah, especially the high resolution results, those don't seem realistic as if the card does have 800 SP or even 480 SP with a shader clock then it will definitely be starving for bandwidth at the really high end. I concur, 4850 results look legit, 4870, not so much

[fng77]

Hmm if those scores are real then 4850 is kinda what I expected but 4870 is only what 10% faster than 4850 at best?! Thats kinda hard to believe.

I also reacted a little to the relativley low difference between the two. But i think we can se the same phenomenon with the 3850 and 3870. I think that an future driver update could help separate them.

[perkam]

Some "new" inq info:

http://www.theinquirer.net/gb/inquir...ormance-posted

The slides on the actual website are old X_X

Perkam

[AliG]

I also reacted a little to the relativley low difference between the two. But i think we can se the same phenomenon with the 3850 and 3870. I think that an future driver update could help separate them.

But the key difference here is that the 3870 had only a faster core clock, the 3850 didn't need the extra bandwidth gddr4 brought, while the 4870 on the other hand probably will make use of the extra bandwidth not to mention the core clock difference is quite large this time (125mhz, or in other words, roughly 25% more than the 4850)

[Helmore]

But the key difference here is that the 3870 had only a faster core clock, the 3850 didn't need the extra bandwidth gddr4 brought, while the 4870 on the other hand probably will make use of the extra bandwidth not to mention the core clock difference is quite large this time (125mhz, or in other words, roughly 25% more than the 4850)

No, that's exactly 20% more.....

EDIT: Here is a whole slew of benchmarks on the 'unnamed card' (4850 probably )

[AliG]

whoops was thinking 125/525

[MrHydes]

AMD and Havok announce plans to optimize physics processing on the AMD platform

SUNNYVALE and SAN FRANCISCO, Calif. — June 12, 2008 — AMD
(NYSE: AMD) and Havok today announced plans to jointly investigate
the optimization of physics effects utilizing AMD’s full line of products.


With over 100 developers and 300 leading titles already using Havok’s
physics engine - Havok Physics – the company has clearly defined its
position as the leading developer of game physics. By working
together, both companies are demonstrating their commitment to open
standards and continued support for the needs of the game
community.


“As the complexity and visual fidelity of video games increases, AMD wants to take advantage of opportunities to improve the game experience,” said Rick Bergman, senior vice president and general manager, Graphics Products Group, AMD. “By working with the clear market leader in physics software, AMD can optimize our platforms to consistently deliver the best possible visual experience to the gamer.”

Source: Pcper

Havok vs nvidia physicx aka ageia

[BenchZowner]

This is weird.
As far as I remember, Intel acquired Havok.

[w0mbat]

Yes, but Havok is still independent.

[Buck47]

An independent, wholly-owned subsidiary of Intel.

[SoulsCollective]

This is weird.
As far as I remember, Intel acquired Havok.

In much the same way as AMD acquired ATi, and ATi continues to license Crossfire to Intel

[perkam]

This is weird.
As far as I remember, Intel acquired Havok.

I dont think you understand the situation. This has nothing to do with AMD, and more to do with Larrabee, ATI and Nvidia

If Intel had not had ATI on board for Havok, Nvidia would have tried to lure them towards their own Physx API, and that would have meant that the greater than 90% of the standalone graphics market would have gone towards Physx, basically rendering Havok useless until the launch of Larrabee in late 2009/early 2010, by which time a greater portion of current and future games would be developed to support Physx

Perkam

[fng77]

But the key difference here is that the 3870 had only a faster core clock, the 3850 didn't need the extra bandwidth gddr4 brought, while the 4870 on the other hand probably will make use of the extra bandwidth not to mention the core clock difference is quite large this time (125mhz, or in other words, roughly 25% more than the 4850)

Yes, you are right. There should indeed be a bigger difference between them.

[AliG]

exactly my point, the 4850 results could be legit, but if they are, the 4870 results are fake, the other is also possible, but if the card does have 800 shaders I doubt it

[BenchZowner]

I dont think you understand the situation. This has nothing to do with AMD, and more to do with Larrabee, ATI and Nvidia

If Intel had not had ATI on board for Havok, Nvidia would have tried to lure them towards their own Physx API, and that would have meant that the greater than 90% of the standalone graphics market would have gone towards Physx, basically rendering Havok useless until the launch of Larrabee in late 2009/early 2010, by which time a greater portion of current and future games would be developed to support Physx

Perkam

You're right.
I wasn't really thinking about it it seems

[flopper]

http://forums.vr-zone.com/showthread.php?t=287874

4850 crossfire result?

[Genius_pt]

If that is true, one word: IMPRESSIVE!

[tajoh111]

http://forums.vr-zone.com/showthread.php?t=287874

4850 crossfire result?

Impressive if true, I had no Idea crossfire scaled that well.

Before people say OH my God its faster than a 280 gtx, look at the cpu frequency. 4.0 ghz.

[bill_d]

Nvidia did say cpu's no longer matter

[003]

Before people say OH my God its faster than a 280 gtx, look at the cpu frequency. 4.0 ghz.

CPU doesn't have nearly as large of an effect on vantage as it did in 06 ... especially in the extreme preset. Wow, and a 4850X2 will be even faster due to the shared memory pool. Nvidia is about to get

I loved how they made that statement about how "you only invest in making a top end video card if you know it will be the best", and here they fail, while ATI probably wasn't even TRYING to make a top end video card!

And finally, the score of the GTX280 of X4800 was on the same CPU as this crossfire setup. Says so in the first post of the linked thread.

[tajoh111]

CPU doesn't have nearly as large of an effect on vantage as it did in 06 ... especially in the extreme preset. Wow, and a 4850X2 will be even faster due to the shared memory pool. Nvidia is about to get

I loved how they made that statement about how "you only invest in making a top end video card if you know it will be the best", and here they fail, while ATI probably wasn't even TRYING to make a top end video card!

And finally, the score of the GTX280 of X4800 was on the same CPU as this crossfire setup. Says so in the first post of the linked thread.

Look at the latest response other person doesn't know what they are talking about.

Also refer to this

http://forums.vr-zone.com/showthread.php?t=287812

3.0 ghz X5000 score

Plus every other review has shown a 3.0 ghz cpu.

This is the first 4.0 I seen in any of these leaks.

[003]

Look at the latest response other person doesn't know what they are talking about.

Also refer to this

http://forums.vr-zone.com/showthread.php?t=287812

3.0 ghz X5000 score

Plus every other review has shown a 3.0 ghz cpu.

This is the first 4.0 I seen in any of these leaks.

Ok. Not much difference though, as vantage is not a CPU benchmark like 06 was (though 06 is a great benchmark on highest graphics settings), and it becomes even less of a CPU test on the extreme preset. If this is just 4850 crossfire, wait for 4870 crossfire and the 4870 X2 .... not even to speak of the dual 4870X2s ..

[fng77]

http://forums.vr-zone.com/showthread.php?t=287874

4850 crossfire result?

Hope so!

[SNiiPE_DoGG]

http://forums.vr-zone.com/showthread.php?t=287874

4850 crossfire result?

now i am so glad i have an x38 board my 2 4870s are gonna rock

[clonez]

looks great

lets see how tri-sli gtx280 fares against quadfire 4870x2