I dont mean to spread fud..

[ChrisRay]

But has this been explained?

I was also reading on the futuremark forums That the Cat 4.5 Betas are faster in 3dmark2003 as well compared to the others, (Most prefer em)

Any idea whats going?


Keep in mind I am asking this because, Not quite sure where it comes from ect, What server as well. And if anyone else can verify the performance differences between 4.5 beta and 4.5 release.

 

[jvd]

The certification takes over a month. So the 4.5betas used in reviews would be a much newer build than the ones released for consumers.

Also the ones released on the cd for the x800pro are no the 4.5s on the website

 

[ChrisRay]

The certification takes over a month. So the 4.5betas used in reviews would be a much newer build than the ones released for consumers.

Also the ones released on the cd for the x800pro are no the 4.5s on the website

So ATI users can expect to see this kind of performance increase in a future release? Okey all I was asking.

 

[jvd]

The certification takes over a month. So the 4.5betas used in reviews would be a much newer build than the ones released for consumers.

Also the ones released on the cd for the x800pro are no the 4.5s on the website

So ATI users can expect to see this kind of performance increase in a future release? Okey all I was asking.

why wouldn't they . They are newer drivers . I'm sure at one point they or a faster set will be put through .

Its really no diffrent than those drivers nvidia gives out before reviews .

 

[KimB]

Except for the simple fact that the X800 Pro is essentially the same basic architecture as the previous R3xx chips, so it seems rather hard to believe that significant performance increases could be had at this stage in the game (at almost two years since the R300 was released). But, I suppose we'll find out sooner or later.

 

[ChrisRay]

The certification takes over a month. So the 4.5betas used in reviews would be a much newer build than the ones released for consumers.

Also the ones released on the cd for the x800pro are no the 4.5s on the website

So ATI users can expect to see this kind of performance increase in a future release? Okey all I was asking.

why wouldn't they . They are newer drivers . I'm sure at one point they or a faster set will be put through .

Its really no diffrent than those drivers nvidia gives out before reviews .

Well. Those are pretty significant performance increases, Usually I expect performance in line with 2% from improved drivers, But if thats the case, the X800 Pro users have alot to look forward too.,

 

[jvd]

Except for the simple fact that the X800 Pro is essentially the same basic architecture as the previous R3xx chips, so it seems rather hard to believe that significant performance increases could be had at this stage in the game (at almost two years since the R300 was released). But, I suppose we'll find out sooner or later.

SIgh Chalnoth we've seen for years drivers from nvidia that apear out of no where and give all cards a huge performance increase . Why wouldn't a new card with new memory controllers and new feature not have alot of power left untapped ?

Why do you think only nvidia is capable of that ?

Well. Those are pretty significant performance increases, Usually I expect performance in line with 2% from improved drivers, But if thats the case, the X800 Pro users have alot to look forward too.,

I normaly expect that too. With cards already out on the market for a long time. But this is more than just a super r3x0 . There are other things that could have been improved that we will never know about. There could also have been broken or bugged hardware that has been fixed.

Not only that but ati is totaly rewriting thier open gl drivers . Perhaps they have done something like this for the direct 3d drivers .

 

[dan2097]

On the pessimistic side, it could be just that the catalyst 4.5s only have basic support for the X800 and the 4.5 betas bring the frame rates up to the levels you'd expect for the X800, thus no nice gains for the r3xx (but we can but hope)

The X800 does still have areas where it could use driver improvements like the memory controller, but these improvements look to be mainly in ps performance, which has got room for improvement, but I would of thought would be more incremental in iuncrease (and this would apply for the whole r3xx/r4xx)

EDIT: has anyone run tests like this on an r3xx?

 

[Althornin]

Except for the simple fact that the X800 Pro is essentially the same basic architecture as the previous R3xx chips, so it seems rather hard to believe that significant performance increases could be had at this stage in the game (at almost two years since the R300 was released). But, I suppose we'll find out sooner or later.

oh whatever. You didnt say that about the GF4 over the GF3, or anything else.
Stop spouting nonsense. There are plenty of areas for optimization, some of which have been discussed by ATI in these very forums.

 

[KimB]

SIgh Chalnoth we've seen for years drivers from nvidia that apear out of no where and give all cards a huge performance increase . Why wouldn't a new card with new memory controllers and new feature not have alot of power left untapped ?

Why do you think only nvidia is capable of that ?

That's not what I'm saying. What I'm saying is that the X800 appears to be too similar to the R3xx architecture to achieve such significant performance increases from "regular" optimizations in new drivers.

 

[KimB]

oh whatever. You didnt say that about the GF4 over the GF3, or anything else.
Stop spouting nonsense. There are plenty of areas for optimization, some of which have been discussed by ATI in these very forums.

Were there ever any drivers for the GeForce4 that significantly improved performance? I really don't remember, so I'm actually asking.

 

[AlphaWolf]

Were there ever any drivers for the GeForce4 that significantly improved performance? I really don't remember, so I'm actually asking.

They released a driver set around the time of the 9700pro that was supposed to have big performance improvements, I don't really recall what the break down was on that.

That's not what I'm saying. What I'm saying is that the X800 appears to be too similar to the R3xx architecture to achieve such significant performance increases from "regular" optimizations in new drivers.

Just because they 'appear' similar to you, it doesn't mean there isn't performance to be found. I believe it has been mentioned that the memory controller is apparently quite different which can have a major impact on some scores.

 

[ChrisRay]

Were there ever any drivers for the GeForce4 that significantly improved performance? I really don't remember, so I'm actually asking.

They released a driver set around the time of the 9700pro that was supposed to have big performance improvements, I don't really recall what the break down was on that.

That's not what I'm saying. What I'm saying is that the X800 appears to be too similar to the R3xx architecture to achieve such significant performance increases from "regular" optimizations in new drivers.

Just because they 'appear' similar to you, it doesn't mean there isn't performance to be found. I believe it has been mentioned that the memory controller is apparently quite different which can have a major impact on some scores.

There was the nature optimisations that went into 3dmark2001 in the 4x.xx dets which dramatically improved scores., These optimisations magically vanished the with 5x.xx dets tho

 

[Ailuros]

oh whatever. You didnt say that about the GF4 over the GF3, or anything else.
Stop spouting nonsense. There are plenty of areas for optimization, some of which have been discussed by ATI in these very forums.

Were there ever any drivers for the GeForce4 that significantly improved performance? I really don't remember, so I'm actually asking.

Yes for 3dmark2k1 mostly. Nothing though in the end that must have been much different with R2xx's too (see Mother Nature performance increases f.e.).

Those type of performance increases are in their majority anyway only for the most popular key applications (yes I know it's common knowledge).

 

[Lezmaka]

The certification takes over a month. So the 4.5betas used in reviews would be a much newer build than the ones released for consumers.

Unless the testing process has changed dramatically in the past year, certification should take nowhere hear a month.

http://www.driverheaven.net/terrymakedon2/

Zardon: Will all future ATI driver releases be WHQL? Do you feel this is necessary, or would you be willing to release drivers on even tighter schedules in the future, if you felt demand warranted this?

Terry: CATALYST is based on the foundation of quality. As such we feel it is imperative that a neutral third party give our drivers a solid run-through. The only authority to do this of course is Microsoft. So yes all our driver releases will be WHQL’ed.

Note: I realize your question is alluding to more frequent postings. If that is the case then not certifying our drivers is not the way to do that. Certification takes us no longer than three days, and we truly believe it is a very important three-day investment. If users feel they want more frequent postings then let us know and we will accommodate. We will NOT however compromise driver quality by not obtaining Microsoft certification.

So either the tests have changed quite a bit since that interview, or he was misinformed or talking about something different, etc.

 

[Evildeus]

Interesting.

 

[Dave Baumann]

Seriously, are we questioning why newer drivers on a newer architecture have more performance than very early ones?

And, no Chalnoth, this isn't exactly the same as R3x0 with more pipelines as there have been many micro architecture and specific changes to the pipeline. There is room for making some significant gains just from a R3x0 driver path.

As for certification, AFAIK the full certification process can take a couple of days to run. The process is actually run by the IHV, so they will know if they are going to get certification by looking at the results. However, AFAIK the process needs to be run for every board the drivers supports - when the process is finished encrypted log files are produced and these are submitted to microsoft for review (if you have some failures you then can haggle a little with MS) and then once MS has reviewed them the IHV is supplied with a WHQL digital certificate. You'll know if a board will be certified withing a few days of running the test - its usually the submittal to MS that takes the time.

 

[ChrisRay]

Seriously, are we questioning why newer drivers on a newer architecture have more performance than very early ones?

And, no Chalnoth, this isn't exactly the same as R3x0 with more pipelines as there have been many micro architecture and specific changes to the pipeline. There is room for making some significant gains just from a R3x0 driver path.

As for certification, AFAIK the full certification process can take a couple of days to run. The process is actually run by the IHV, so they will know if they are going to get certification by looking at the results. However, AFAIK the process needs to be run for every board the drivers supports - when the process is finished encrypted log files are produced and these are submitted to microsoft for review (if you have some failures you then can haggle a little with MS) and then once MS has reviewed them the IHV is supplied with a WHQL digital certificate. You'll know if a board will be certified withing a few days of running the test - its usually the submittal to MS that takes the time.

I think we're questioning, Why the review drivers show so much more performance than the official Ones, And if these "performance" boosts will come in a future release, Maybe 4.6

 

[Dave Baumann]

I would imagine so. As I mentioned before - FM tested these and were OK a far as 3DMark03 were concerned, so if the PS gains were legit there it would probably indicate that this is generic hence the gains in other PS titles.

 

[ChrisRay]

I would imagine so. As I mentioned before - FM tested these and were OK a far as 3DMark03 were concerned, so if the PS gains were legit there it would probably indicate that this is generic hence the gains in other PS titles.

Just Curious, Have there been any comparisons really made between the Beta 4.5 and the Normal 4.5s? It would be nice for 4.6 to hurry up and get here so we can see, If I had an X800 Pro I'd be doing all sorts of comparisons.

*loves ne hardware* ;p