Friday, February 20th 2009

ATI Catalyst 9.2 Released

AMD has just released ATI Catalyst version 9.2, so now you have a brand new driver for tonight and a nice tweaking utility to test. Catalyst 9.2 brings performance benefits in several games:
  • Crysis DX10 gains up to 20%
  • Crysis Warhead DX10 gains up to 20%
  • World in Conflict gains up to 5%
Now the new driver also enables Anisotropic Filtering for DirectX 10 applications in Windows Vista and fixes some minor issues that occured in previous versions. For the full list of fixes, please read the Catalyst 9.2 release notes.

DOWNLOAD: ATI Catalyst version 9.2 for Windows XP and Windows Vista (32/64-bit versions)Source: AMD Game!
Add your own comment

174 Comments on ATI Catalyst 9.2 Released

#1
Wile E
Power User
niko084 said:
Well, that would be my initial thoughts too, but this is a "driver crash", and again only with the "9.2 drivers", with the "9.1 drivers" everything is perfectly fine. Also Crysis is the only program it has issues with, including numerous stress tests, benchmarks and other games for extended periods of time.

Now ruling out that it is not a perfectly stable clock, but it doesn't seem to fit all in properly. No artifacts, no excessive heat, every other piece of software runs perfectly well.
The point still stands, tho. If it kept crashing, even with lowering the clock, it would purely be a driver issue, but since lowering the clocks fixes it, it's instability, no matter how you look at it. The new drivers may just be more sensitive.

Either way, 20Mhz isn't going to make a real performance difference anyway.
Posted on Reply
#2
niko084
Wile E said:
The point still stands, tho. If it kept crashing, even with lowering the clock, it would purely be a driver issue, but since lowering the clocks fixes it, it's instability, no matter how you look at it. The new drivers may just be more sensitive.

Either way, 20Mhz isn't going to make a real performance difference anyway.
Well I wont argue the point on the 20mhz either way for sure.

What I don't really understand is how an overclock on the card can cause driver failure and only in a particular application. It's not something I have ever seen before, not a video card blowing the driver.
Posted on Reply
#3
Super XP
Silverel said:
More than enough information here to show that there are indeed large changes between dx9 and dx10. Whether or not games use it is their problem.

http://www.hardwaresecrets.com/article/95/2


Thanks for the info, I was on MS's website looking for such information. So there is a large difference between the two. I wonder what will DX11 bring to the table, seeing that DX10 has a lot to offer and we havent even tapped into it fully just yet.
Posted on Reply
#4
Super XP
Interesting!
DirectX explained:
In general a new version is launched to permit the system to recognize the most modern features of the graphic chips, so it can use these features in case your video card have them, increasing both the 3D performance of the machine and the image quality.

DirectX 10 brought Shader 4.0 model. This model brought an important change in the architecture of graphics chips. Up to DirectX 9.0c-based graphics chips, the chip has separated processing units for processing pixel shaders and vertex shaders. In some circumstances all pixel shader units were being used, the vertex shader units were idle and new pixel shader instructions needed to wait, because all pixel shader processors were busy, even though the vertex shader processors were idle. On DirectX 10-based graphics chips the processing units are generic and can be used to process any kind of information, solving the problem we’ve just explained: on DirectX 10-based video cards instructions need to wait only when all processors are busy, regardless of the kind of instruction. This architecture is called Unified Shader Architecture. DirectX 10 also introduced a new shader type, geometry, and other features to transfer more processing from the system CPU to the graphics chip.
Posted on Reply
#5
kiriakost
Super XP said:
Interesting!
:laugh: :laugh: :laugh: Now you know why i replaced the 6800GT with the HD3850 ;)

I care more for the hardware changes , than DX it self .
Posted on Reply
#6
Drizzt5
I don't understand why the cards out now can't support dx11 when it comes out.
Posted on Reply
#7
niko084
Drizzt5 said:
I don't understand why the cards out now can't support dx11 when it comes out.
They maybe able to "technically" but also they may require new "hardware" like Dx10 needs Shader V 4.0...
Posted on Reply
#8
Silverel
kiriakost said:
:laugh: :laugh: :laugh: Now you know why i replaced the 6800GT with the HD3850 ;)

I care more for the hardware changes , than DX it self .
DX10 brought about the changes in the hardware requirements as above. Without one, there wouldn't have been another, and we would be at the limit on hardware for DX9 already.
Posted on Reply
#9
Super XP
I think DX11 will be based on multi-GPU graphics cards. ATI should be releasing a 8 to 10+ GPU graphics card. Not sure if its the RV880 or the RV980.
Posted on Reply
#10
kiriakost
Silverel said:
DX10 brought about the changes in the hardware requirements as above. Without one, there wouldn't have been another, and we would be at the limit on hardware for DX9 already.
Not true ... or half true ....

Many years back , the hardware had to offer support in Three widespread protocols.
DX + Open GL + Glide ( 3dfx) .

Now days that .. DX survived ( do not like to get more to it, of how it did that) .. everything are easier for the hardware makers .
I disagree that the DX development pushes ahead the hardware development .

Its all about cost in the bottom line , and the hardware makers do mostly was is cheap to deal with.

In the museum of " software hypes that never offered anything " ...
they are many exhibits ... like MMX & shader models , and many more . ;)
Posted on Reply
#11
Super XP
kiriakost said:
Not true ... or half true ....

Many years back , the hardware had to offer support in Three widespread protocols.
DX + Open GL + Glide ( 3dfx) .

Now days that .. DX survived ( do not like to get more to it, of how it did that) .. everything are easier for the hardware makers .
I disagree that the DX development pushes ahead the hardware development .

Its all about cost in the bottom line , and the hardware makers do mostly was is cheap to deal with.

In the museum of " software hypes that never offered anything " ...
they are many exhibits ... like MMX & shader models , and many more . ;)
Did you work for MS? ;)
Where from Greece are you from? I have loads of family in KOS and Tripoli louka.
Posted on Reply
#12
kiriakost
Super XP said:
Did you work for MS? ;)
:laugh: i can only say that i have study , one certification program MCSE 2000 server.
But its also a hobby of my , or passion , to study and find the truth over the marketing hypes.


Super XP said:

Where from Greece are you from? I have loads of family in KOS and Tripoli louka.
I am from Thessaly , and the city of Volos , the land of Jason and the argonauts . ;)
Posted on Reply
#13
Silverel
kiriakost said:
Not true ... or half true ....

Many years back , the hardware had to offer support in Three widespread protocols.
DX + Open GL + Glide ( 3dfx) .

Now days that .. DX survived ( do not like to get more to it, of how it did that) .. everything are easier for the hardware makers .
I disagree that the DX development pushes ahead the hardware development .

Its all about cost in the bottom line , and the hardware makers do mostly was is cheap to deal with.

In the museum of " software hypes that never offered anything " ...
they are many exhibits ... like MMX & shader models , and many more . ;)
That's an opinion in face of the truth. Without DX10 forcing new hardware requirements, hardware would not have a reason to advance. It is all about cost in the bottom line after all. Hardware vendors make their money selling the easiest stuff to produce over and over again with very limited changes. Software vendors make their money by selling the dynamic stuff that requires hardware changes, or adds essential (and often 'required') functionality.

Considering that DX leads the pack in Gaming development by a large margin, it's calling the shots on hardware requirements in the form of DX releases. OpenGL is a minor factor for a handful of niche gaming developers, and Glide is dead.

To dig in even further to prove the fact, take a look at GPU benchmarks. The 1950pro is a DX9.0c card from many years ago. Even to this day it can put up ~90fps in CoH at 16x12 resolution. This is the equivalent of a 9600GSO, or HD3850. Even though the hardware has "significantly advanced", a generations old card can still compete in the DX9 arena. Simply put, because DX9 has not changed, therefore hardware has not changed.

Point being, you cannot run a 1950pro in a DX10 game. The closest you can get is emulating DX10 effects in DX9 mode. It simply does not have the HARDWARE capabilities.
Posted on Reply
#14
kiriakost
Silverel said:

Point being, you cannot run a 1950pro in a DX10 game. The closest you can get is emulating DX10 effects in DX9 mode. It simply does not have the HARDWARE capabilities.
I love quality conversations ;) .. in a world loaded with dual cores , and 70-200FPS coming from fresh hardware .

I like to believe that few shaders , even emulated by software, it will not force any system to fall at it knees .

Ultimate performance = latest only goods .
High performance = something that most can have .
Low performance = truly incompatible hardware. ( VGA that does not support DX9 )
Posted on Reply
#15
Wile E
Power User
kiriakost said:
I love quality conversations ;) .. in a world loaded with dual cores , and 70-200FPS coming from fresh hardware .

I like to believe that few shaders , even emulated by software, it will not force any system to fall at it knees .

Ultimate performance = latest only goods .
High performance = something that most can have .
Low performance = truly incompatible hardware. ( VGA that does not support DX9 )
But DX9 cards do fall to their knees trying to run DX10 effects. I think that is the point Silverel is trying to make. DX9 cards simply cannot do SM4 instructions without suffering huge performance penalties.

Therefore, new hardware was needed. That came in the way of the 8800 series with SM4 capabilities.
Posted on Reply
#16
niko084
Wile E said:
But DX9 cards do fall to their knees trying to run DX10 effects. I think that is the point Silverel is trying to make. DX9 card simply cannot do SM4 instructions.

Therefore, new hardware was needed. That came in the way of the 8800 series with SM4 capabilities.
A DX9 card running Dx10 effects would have to run it all on the gpu woudln't it?
Posted on Reply
#17
Wile E
Power User
niko084 said:
A DX9 card running Dx10 effects would have to run it all on the gpu woudln't it?
You could probably try to do cpu emulation, but the performance hit would still be ridiculous.

When DX10 first released, and all those tech demos came out, people hacked them to be able to run on DX9 hardware. Not even the fastest DX9 cards could pull it off without it being a slide show.
Posted on Reply
#18
niko084
Wile E said:
You could probably try to do cpu emulation, but the performance hit would still be ridiculous.

When DX10 first released, and all those tech demos came out, people hacked them to be able to run on DX9 hardware. Not even the fastest DX9 cards could pull it off without it being a slide show.
Ya, thats what I figured, I wouldn't doubt the slide show, you can beat hardware with software emulation, it just doesn't work... Maybe if you were doing software emulation of Dx8 on new Core2's and I7's but we are talking about a pretty large generation step there.
Posted on Reply
#19
kiriakost
Wile E said:

When DX10 first released, and all those tech demos came out, people hacked them to be able to run on DX9 hardware. Not even the fastest DX9 cards could pull it off without it being a slide show.
Can i have a link please ... :)
Posted on Reply
#20
Wile E
Power User
kiriakost said:
Can i have a link please ... :)
lol. That was years ago. Good luck digging up those links. Iirc, there were posts in this forum about it. Happy hunting. lol.
Posted on Reply
#21
Super XP
kiriakost said:
:laugh: i can only say that i have study , one certification program MCSE 2000 server.
But its also a hobby of my , or passion , to study and find the truth over the marketing hypes.


I am from Thessaly , and the city of Volos , the land of Jason and the argonauts . ;)
Oh, Mythology, my favorite which I do believe was for real. I've taken 2 years of Ancient Greek to learn how to read the holly scriptures and mythological readings which were written in Ancient Greek Literature.

I sometimes get people talking about Socrates of KOS Greece, the first Doctor. If my memory serves me well, his teachings and principals are still used today in the School of Medicine and Doctoring.
Posted on Reply
#23
Super XP
Almost every single aspect in regards to Greek Mythology, is based on truth. It happened, just like the sirens, Hercules, Poseidon, Troy, Alexander the Great, Arcilies, the golden fleece etc, what man could not explain back then can be explained with "tomorrows" physics & technology.
Posted on Reply
#24
Azkeyz
1. Get some memtest86
2. Burn it to a CDR

boot up the computer with the memtest86 cd

If you get a failure on the first pass you could have a bad memory module or just need to clean and reseat the memory.

You can wipe the memory pins clean with a paper towel and some cellulitis ridden cat saliva

Kidding, use human saliva ...
Posted on Reply
Add your own comment