Catalyst drivers are "FAIL"? Please elaborate!

[Wrigleyvillain]

So what does MLAA really bring to the table in people's opinions? I haven't yet done my homework on it besides a cursory glance at some screenshots.

 

[jagd]

Truth tells the opposite , MS OS crash report showed between ati and nvida nvidia drivers cause much more problem ( if you want to see numbers google it).This phrase is part of nvidia's "social media marketing " scheme i think
Edit :found link http://arstechnica.com/hardware/new...it-paints-picture-of-buggy-nvidia-drivers.ars
28.8% nvidia vs 9.3 ATI

Ok, look, I see this phrase thrown around a lot. There is definitely an impression that AMD has great hardware but awful drivers compared to those from Nvidia and it is vocalized often.

 

[dir_d]

So what does MLAA really bring to the table in people's opinions? I haven't yet done my homework on it besides a cursory glance at some screenshots.

So far it really helps older games that dont have a type of AA setting. It also seems to not take a big hit and is comparable to 8x MSAA. MLAA can only be as good as the drivers though so it knows exactly what to process and what not to. People have put MLAA in a negitive light because they say it blurs textures but i dont see it when running on my own hardware.

 

[Wrigleyvillain]

Thanks. One of the screen shots I saw was Dead Space and it looked pretty sharp with MLAA compared to no-AA vanilla (which was one of the reasons this is still sitting in my Steam library unplayed (along with it's kinda wonky controls)).

 

[TheoneandonlyMrK]

imho im ALLWAYS aiming for better looking gfx so i would never need to disable AA or antisotropy quite the opposite and i think as some do that both nv and amd have issues from time to time and in my oppinion its best to go with whichever is selling the best stuff when your buying.
drivers mean less with time anyway as ati tray tools and nv tools and the like allow you to force options in game so only stability then matters and MLAA is effin fantastic ive had many diff nv and ati cards and its honestly the best driver bound improvement ive yet seen implemented on my pc it seems to just work v well 4x multi on also and i max every game still and still get v playable fps but id say hybrid physx helps maintiain that point in games that use it

 

[TheoneandonlyMrK]

and also ive an gt240nv and a 5870 in my pc attached to the same 32"sony bravia and at the same res the ati card looks miles nicer nv by comparrison imho looks a bit XBOX resy a deffinate over flushed with colour look bit dark and slightly grainy looking to be specific

 

[Wrigleyvillain]

^ Thanks. I notice the same thing in terms of general superior IQ with AMD though I wouldn't go so far as to call NV "grainy". Though at least for DX11 games it makes sense AMD would have an advantage as they had DX11 hardware available so much earlier than nv.

 

[TheoneandonlyMrK]

i agree grainy might snd harsh and i dont mean grainy it i kinda ment it looks like er i dunno if i got close to the screen it would look grainyer farrer away ,just less detailed or a bit muted

 

[TheoneandonlyMrK]

and so far as driver issues iv had efffin loads but thats born out of the fact i keep upto date with drivers but even then games come out that catch ati/nvidia out a bit then play a bit poo like cod black ops it shows many issues in its log some of which are driver related maybe micro stuttering randomly being one but akin to lag?
i had most issues with dual 3870s when they had just come out but after a year they were comparitively amazing faster more stable and higher fps in all games it just takes them both a bit of time to optimise drivers right i had issues with my prophet gts and gt240 inc all the 260 drivers which crash on install every time then mess my gfx card up so its not seen not updated since 158 just the physx niether are perfect both a little bit shit

 

[LAN_deRf_HA]

I know nvidia and ati have about the same amount of officially documented bugs, but I've only ever run into the ati ones. I can only think that has something to do with what you're using it for. Like looking at the nvidia bugs I see a unigine bug when run on a projector, or when you're rotating your display on hdmi. Stuff I'd never do. Looking at the ati list I see a lot of game issues, and it occurs to me the root of that might have something to do with the games themselves. Being better tested on or sponsored by nvidia. I think because of games being built for nvidia the ati driver team has to do a lot more work.

 

[Mussels]

So this is the same as Nvidia's "enhance" in-game AA setting I am assuming.

MSAA + SSAA = superduper awesome AA!

 

[Wrigleyvillain]

^ I can't tell if you are joking or not lol

 

[alexsubri]

So what does MLAA really bring to the table in people's opinions? I haven't yet done my homework on it besides a cursory glance at some screenshots.

My video
Here are the pics all taking with 16x MLAA [56k warning]

 

[newtekie1]

Truth tells the opposite , MS OS crash report showed between ati and nvida nvidia drivers cause much more problem ( if you want to see numbers google it).This phrase is part of nvidia's "social media marketing " scheme i think
Edit :found link http://arstechnica.com/hardware/new...it-paints-picture-of-buggy-nvidia-drivers.ars
28.8% nvidia vs 9.3 ATI

When these numbers came out everyone pretty much ignored them because they don't take into account that nVidia releases a lot of Beta drivers to the public causing a lot more crashes. If only the nVidia drivers that were WHQL certified were counted the number would likely be closer to ATi's.

 

[jagd]

I understand you but we are talking about " good nvidia drivers myth" .These numbers may caused by beta drivers but it does not change end result ,nvidia could release good working WHQL certifed drivers, who forced them to release betas? Beta drivers are released most of the time to fix problems caused by another driver version does not it?

When these numbers came out everyone pretty much ignored them because they don't take into account that nVidia releases a lot of Beta drivers to the public causing a lot more crashes. If only the nVidia drivers that were WHQL certified were counted the number would likely be closer to ATi's.

 

[newtekie1]

I understand you but we are talking about " good nvidia drivers myth" .These numbers may caused by beta drivers but it does not change end result ,nvidia could release good working WHQL certifed drivers, who forced them to release betas? Beta drivers are released most of the time to fix problems caused by another driver version does not it?

No, not really. The WHQL drivers tend to be bug fixes, the betas tend to be performance improvements. I can't remember the last bug fix that came from a beta drivers. Releasing betas is a good thing because it means their WHQL drivers tend to be better than ATi's, and they don't need multiple hot fixes. No one is forcing anyone to use the beta drivers.

And the crashes caused by beta drivers shouldn't count because you have to go to a special section to get them, they aren't offered to the normal user just searching for drivers, and it is understood that they may be less stable.

 

[wahdangun]

No, not really. The WHQL drivers tend to be bug fixes, the betas tend to be performance improvements. I can't remember the last bug fix that came from a beta drivers. Releasing betas is a good thing because it means their WHQL drivers tend to be better than ATi's, and they don't need multiple hot fixes. No one is forcing anyone to use the beta drivers.

And the crashes caused by beta drivers shouldn't count because you have to go to a special section to get them, they aren't offered to the normal user just searching for drivers, and it is understood that they may be less stable.

nah i think they are not better than ati, remember nvdia driver disable the fan and make the card overheating.

and btw GTX460 has only had 3 driver releases since July....and nVidia is taking their sweet time fixing major issues. The image quality has even gone downhill.... tried to watch a blu-ray movie the other night. Was like watching an 80's era CRT TV....pixelated all to hell.

 

[Solaris17]

Honestly iv seen driver bugs/bad drivers from both sides. but it really comes down to the games I play. The ones I play simply run better on nvidia cards/ and thats apparent in the FPS I get. I do however absolutely know that ATI's crossfire blows. I have never seen such terrible scaling issues. But that might come down to architecture so meh.

Truth tells the opposite , MS OS crash report showed between ati and nvida nvidia drivers cause much more problem ( if you want to see numbers google it).This phrase is part of nvidia's "social media marketing " scheme i think
Edit :found link http://arstechnica.com/hardware/new...it-paints-picture-of-buggy-nvidia-drivers.ars
28.8% nvidia vs 9.3 ATI

thats not a fair comparioson at all. nvidia's drivers are ment to throw an error and reset when they are overclocked to far etc. I dont remember that being an option with ATI.

 

[Steevo]

Honestly iv seen driver bugs/bad drivers from both sides. but it really comes down to the games I play. The ones I play simply run better on nvidia cards/ and thats apparent in the FPS I get. I do however absolutely know that ATI's crossfire blows. I have never seen such terrible scaling issues. But that might come down to architecture so meh.



thats not a fair comparioson at all. nvidia's drivers are ment to throw an error and reset when they are overclocked to far etc. I dont remember that being an option with ATI.

My driver recovers when I get stupid. Even when I try and run the GPU at 1150 instead of 1050, ATI driver recovery has been there for years.

Last night Test Drive Unlimited being the piece of shit it is, crashed again. However I was able to end the process and even though the driver was implicated in the crash the fault is Atardi made a piss poor game port that leaks and consumes all the available memory, I was back to the desktop like new.

 

[Radical_Edward]

I've yet to have a driver issue that wasn't due to my own error.

 

[Solaris17]

My driver recovers when I get stupid. Even when I try and run the GPU at 1150 instead of 1050, ATI driver recovery has been there for years.

ok so my bad. but now take into consideration how many people have nvidia over ATI and tell me thats a fair statistic.

lets take a look at the steam survey just for shits and giggles.

http://store.steampowered.com/hwsurvey/

thats just the people that use steam.

now imagine all the other gamers.

ok now imagine how popular netbooks etc are getting

now take into consideration nvidia has contracts with most of them.

are we still fair or...?

 

[Steevo]

ok so my bad. but now take into consideration how many people have nvidia over ATI and tell me thats a fair statistic.

lets take a look at the steam survey just for shits and giggles.

http://store.steampowered.com/hwsurvey/

thats just the people that use steam.

now imagine all the other gamers.

ok now imagine how popular netbooks etc are getting

now take into consideration nvidia has contracts with most of them.

are we still fair or...?

No, as the ratio is off for how many cards were truly in use. Most users have integrated Intel shit, and at that time it was 2:1 Nvidia over ATI on dedicated cards.


So they had almost 33% more crashes per instance than did ATI. We did the math when the report was released.

 

[Solaris17]

No, as the ratio is off for how many cards were truly in use. Most users have integrated Intel shit, and at that time it was 2:1 Nvidia over ATI on dedicated cards.


So they had almost 33% more crashes per instance than did ATI. We did the math when the report was released.

so this isnt a "lol windows has more viruses then mac does" argument how?

 

[Steevo]

When 100 people have cards, 66% have Nvidia, and 30% of those have driver crashes on a regular basis VS the 33% that have ATI cards and only 10% of those user have crashes we arrive at a basic truth.


Numbers don't lie.

30% of 66 users is 20 users rounded
10% of 33 users is 3 rounded.


If you had a car and it had a 30% chance of randomly stalling and another car with a 10% chance, which would you buy?


As far as your Mac VS PC, it doesn't apply. We aren't talking about adoption rates, we are using standardized error rates.

 

[Solaris17]

When 100 people have cards, 66% have Nvidia, and 30% of those have driver crashes on a regular basis VS the 33% that have ATI cards and only 10% of those user have crashes we arrive at a basic truth.


Numbers don't lie.

30% of 66 users is 20 users rounded
10% of 33 users is 3 rounded.


If you had a car and it had a 30% chance of randomly stalling and another car with a 10% chance, which would you buy?


As far as your Mac VS PC, it doesn't apply. We aren't talking about adoption rates, we are using standardized error rates.

ahh i understand now your taking the failure rates of the given percentage.