Are drivers the issue afterall?

Are these the so-called "IQ trade-off" drivers we've heard tell of?
 
We all know the old saying about first impressions. If the problems were mainly driver related, way not wait another couple of weeks and get them striaght.

As is stands now, the 2900's got a lot of negative digital ink that could have been mitigated with better performance, even with the power issues.

That said, you'll get some performance improvments of course with better drivers, but I doubt anything drastic across the board.
 
0mega said:
It's Ragezone. you aren't honestly expecting a lack of bias are you?
Click to expand...

Did you read the thread ? Before and after benches of a driver update 8.36 to 8.37 ?

Looks legit to me. Lots of reviews commented on the rawness of the drivers. Seeing what happend with R520 leads me to believe there will be substantial increases in performance over the next few driver releases.
 
phide said:
Are these the so-called "IQ trade-off" drivers we've heard tell of?
Click to expand...
Well apparently the drivers define the AA algorithms which in turn run on those same 320 shaders that are used for rendering. So you can see that there. So prudent selection of AA type and level is important. Because drivers and even applications can define new AA algorithms the flexibility.

Sort of like giving lots of rope to users. And we all know what getting lots of rope can lead to .... x_X


It is also very disconcerting to see backsliding of performance going from version to version. :(


EDIT: Or are you talking about when a driver gains a defect that actually causes improved performance because something stops getting rendered?
 
All I've heard is the rumour that a driver set exists that gives a massive performance increase whilst at the same time diminishing image quality (how severely I couldn't say). I suppose the large performance differentials between these seemingly minor driver revisions are due to AMD fudging with some number of variables that degrade IQ to some degree.

In any case, it is somewhat curious.
 
phide said:
I suppose the large performance differentials between these seemingly minor driver revisions are due to AMD fudging with some number of variables that degrade IQ to some degree.
Click to expand...
No need to put down to malice what can easily be explained by inneptitude. :p
 
Dwight said:
No need to put down to malice what can easily be explained by inneptitude. :p
Click to expand...

:confused: @ image quality rumors that mean so much to some.
Some are not reading the driver numbers correctly. But its going to be funny when the truth comes out. I wonder how many read the tech repository review :rolleyes:
 
even if drivers were not the issue....the issue of powerdraw and noise...WILL become issues. This card is just plagued!
 
Has everyone forgotten about DX9's HLSL shader compilation, etc. Is it possible to believe that DX9 needs an update to optimize and implement R900? I mean lets be honest here there have been 3 DX 9.0c HLSL shader compilation updates since the release of the G80. I hope we did not forget about this. The way the HLSL shader compilation works could also contribute to the problems as well.
http://msdn2.microsoft.com/en-us/xna/aa937788.aspx
 
You must log in or register to reply here.
Back
Top