The AMD Driver Bug Thread

If I try the #6 fix could that possible help fix my grey screens of death while gaming? My card has crashed also when just sitting in Windows, so I thought I would try.
 
Was never tested as a fix for GSoD issues, but give it a try. It can't curt anything.

Sorry to say, I've heard from a couple of places that if you're still seeing persistent GSoDs at this point, it's likely a hardware problem. You may need to RMA the card.
 
The cursor lag bug is still a huge annoyance for me. Hopefully they get it together soon.
 
Does anyone with the cursor corruption or lag have overscan enabled in Catalyst? I just noticed one time that when I was messing around with mine that it would double in size only when I had the cursor over the little screen bouncing back and forth.

Just thought I'd comment.
 
The cursor lag bug is still a huge annoyance for me. Hopefully they get it together soon.

same here on the #8 cursor lag bug. that and i can't get my secondary display to be detected. not sure what is causing my latter issue though.
 
With regards to #2, I am having an interesting issue that may be related. I have been having issues for the last 3-4 months while watching flash videos. Either on Youtube, MLBTV, or hulu, while a video is playing, my screen turns one color and the sound from the video keeps playing for a second. Then the monitor shuts off and then restarts. I get a notification that my ATI driver stopped responding and needed to be restarted.

After disabling flash hardware acceleration, the issue goes away. Would the fix for #2 also fix this?
 
With regards to #2, I am having an interesting issue that may be related. I have been having issues for the last 3-4 months while watching flash videos. Either on Youtube, MLBTV, or hulu, while a video is playing, my screen turns one color and the sound from the video keeps playing for a second. Then the monitor shuts off and then restarts. I get a notification that my ATI driver stopped responding and needed to be restarted.

After disabling flash hardware acceleration, the issue goes away. Would the fix for #2 also fix this?

Yes, that's related (changing memory clockspeed can make the card go unstable, in addition to causing flickering).

The work-around for #2 should clear up the crashing for you :)
 
Sorry to report, cursor corruption has returned as of Catalyst 11.2. First post has been updated to reflect this regression...

Now we have the cursor corruption bug AND a laggy pointer in the top and bottom right-hand corners. One step forward, two steps back. Get it together AMD, this is getting ridiculous...

Edit: I just launched an Executive Email Carpet Bomb on AMD. Worded a 4-page letter and CCed it to every AMD executive email address I could find (about 30 people in total, including the CEO and the Director). Got a call exactly 3 hours later, directly from the General Manager of AMD's video card division acknowledging that I had stirred up the hive a bit. He's going to call m back tomorrow after he's talked with the software people.
 
Last edited:
I cant believe this cursor problem still is an Issue after all this time. Really?
 
Sorry to report, cursor corruption has returned as of Catalyst 11.2. First post has been updated to reflect this regression...

Now we have the cursor corruption bug AND a laggy pointer in the top and bottom right-hand corners. One step forward, two steps back. Get it together AMD, this is getting ridiculous...

Edit: I just launched an Executive Email Carpet Bomb on AMD. Worded a 4-page letter and CCed it to every AMD executive email address I could find (about 30 people in total, including the CEO and the Director). Got a call exactly 3 hours later, directly from the General Manager of AMD's video card division acknowledging that I had stirred up the hive a bit. He's going to call m back tomorrow after he's talked with the software people.

Good for you (no, I'm being quite serious). I did the same thing with ASUS a while ago and got them to fess up something was broken - no, I won't get back in to it here, they did right by me in the end.

That damn cursor corruption issue has been driving me nuts and I've had zero reply from people I know there.
 
Well, it appears I'm making progress. AMD has been very receptive so far :)

1st call, General Manager confirming they had received my email.
2nd call, General Manager informing me to expect a call from Customer Relations.
3rd call, Customer Relations Supervisor called immediately after.

Spent a good couple hours talking his ear off, going over every issue in this thread in excruciating detail. He's going to call back tomorrow with an engineer or developer once they've had a chance to review the fairly massive list I just dumped on them.

He admitted that this is, to his knowledge, the first time some of the issues in this thread have been brought to their attention (no specifics, though).

Oh, and the cursor corruption bug is back in their bug tracker with a resolution tied to it. Should be fixed FULLY (no more corruption, no more unresponsive corners) in next-months driver.
 
Well, it appears I'm making progress. AMD has been very receptive so far :)

1st call, General Manager confirming they had received my email.
2nd call, General Manager informing me to expect a call from Customer Relations.
3rd call, Customer Relations Supervisor called immediately after.

Spent a good couple hours talking his ear off, going over every issue in this thread in excruciating detail. He's going to call back tomorrow with an engineer or developer once they've had a chance to review the fairly massive list I just dumped on them.

He admitted that this is, to his knowledge, the first time some of the issues in this thread have been brought to their attention (no specifics, though).

Oh, and the cursor corruption bug is back in their bug tracker with a resolution tied to it. Should be fixed FULLY (no more corruption, no more unresponsive corners) in next-months driver.


That's good... now we just have to wait and see if they're serious about it.
Did you tell them about the 2D clock issue with multiple monitors as well?
 
Not sure if it has been mentioned but bezel compensation is only available to identical monitors (model, make and everything), might be a good idea to mention that.
 
Did you tell them about the 2D clock issue with multiple monitors as well?

I mentioned a number of idle clockspeed and PowerPlay issues, nothing specific to multiple monitors though.

Mind detailing what you're talking about so I can test it and relay to AMD?

Not sure if it has been mentioned but bezel compensation is only available to identical monitors (model, make and everything), might be a good idea to mention that.
I think that one might be by design, but I'll mention it.

You can already make bezel correction work on mismatched displays if you clone their EDID's, though.
 
I mentioned a number of idle clockspeed and PowerPlay issues, nothing specific to multiple monitors though.

Mind detailing what you're talking about so I can test it and relay to AMD?

I'll use my HD 5850 as an example:
Overclocking the video card with two (or more) monitors will make the card idle at 157/300 (2D clocks for single monitor) instead of 400/1000 (2D clocks for multiple monitors) --> flickering

Previously it even happened without overclocking when using some driver versions, for all I know it might very well return (as it did in the past). I won't be able to confirm it because I got tired of all the issues and simply edited the 2D clock values to 400/1000 in the card's BIOS.
 
Ok, I did tell him about that one. I told him it happened in both single and dual (or more) monitor configurations, as I've seen it happen to both.

I also mentioned that the idle clockspeed problem keeps popping up, fixed for months only to return again at random in a new driver revision.
 
Really stupid question but are we so sure that AMD is aware of this mouse lag problem?
 
Imade it quite clear that their previous fix had made the issue worse rather than better in my initial email. From my conversations with them, it sounds like they really think it's fixed this time :)
AMD: "The issues revolving around cursor corruption are in the ticketing system and already have a PR attached to them. The problem should be completely fixed"
.
Me: "That includes the unresponsive or laggy pointer we've been seeing in the top-right corner of every display, correct?"
.
AMD: "Yes, the issue should be COMPLETELY resolved now"
They called me back today to get a hardware profile of my system so they can replicate it on their end. They also offered to fly my PC out to headquarters, but man, I can't be without a machine that long.

They don't think the issues I reported are anything specific to my motherboard or other hardware (I gave them a fairly long list of configurations where these issues have appeared), but they still want as close a testbed as possible on their end to check against. Sounds like they basically want an Everest report, as well as information on any weird hardware that Everest doesn't pick up.

They seemed oddly concerned that this was an Asus HD 5850, even after I assured them that it was a reference model from one of the original production runs and not one of Asus's "cost reduced" board revisions. It's even running the AMD reference video BIOS, not the Asus-modified video BIOS. I wonder if I should read anything into that...
 
Awesome Unknown, I appreciate the hard work!

Now let us hold our hands together and hope these bugs are abolished once and for all! :)
 
Heh, just had to spam the company. No big >_>

Seriously though, after finally getting AMD's attention, they've been extremely cooperative. I've had calls from General Managers, a bunch from the PR department, and even a conference call with one of their engineers. They sound determined to straighten this mess out, and I really appreciate it.

I did slip in one tiny feature request while I was at it. I asked them to add a counter to the Eyefinity bezel correction calibration screen that tallies up the number of pixels that bezel correction is going to add to the width of your resolution.

If you've ever tried to re-create an Eyefinity display group with exactly the same resolution you had last time (so you don't have to reconfigure all your games), then you know exactly how difficult it is to tease the calibration tool into giving you a specific resolution. I usually have to go through calibration 5 or 6 times, poking one way or the other until I finally zero in on 5292x1050. The counter should mean we get the resolution we want the first time, every time.
 
Yeah, I thought about that after the fact. I'll wait for some of the other fixes to roll in before piling more stuff on the poor guys.

The default video settings can, at least, be easily changed from within CCC.
 
are all those bugs also affects the 6xxx series in 11.3? Because i'm planning to buy the sapp 6950 next week.
 
I've heard mixed reports on the presence of these issues on the 6800 and 6900 series cards. I don't have one myself, so unfortunately, I can't personally verify their status.
 
I can say something, the powerplay issue really bothered me on my current 6850 direct CU (first revision)... mainly because the original "standard" settings are 100mhz core and 300mhz mem, so yeah due to it being bios based i would say that at least some issues do crop up on the 6800's.

(Fix was to do a profile with 2d clocks of 400mhz core and 1k mem, that way the jump to 800mhz core and 1.15k mem doesn't choke my pc, instantly making my pc stable)
 
Anyone know when the hell they going to fix the mouse cursor lag problem?
It's annoying the fuck out of me....
 
They're building a replica of my system as we speak, I'll let you know what they come up with once they start testing on it.
 
Anyone know when the hell they going to fix the mouse cursor lag problem?
It's annoying the fuck out of me....

Not sure if this works for you, but try turning off Aero Glass. I turned it off for another reason and it appears to be gone. I had a mild case of cursor lag in the upper right corner before.
 
Turning off Aero Glass fixes it for the same reason that opening the Windows Magnifier (temporarily) fixes it.

It disables hardware acceleration on the mouse cursor, and moves the buffer from video RAM (where it was corrupted by the video card or display driver) into system RAM (where Windows can render it in software).
 
They're building a replica of my system as we speak, I'll let you know what they come up with once they start testing on it.

I'm surprised they decided to build a replica, thought maybe since they wanted to fly yours out there, they just would've offered you a sweet AMD rig in exchange for yours.
 
I'm surprised they decided to build a replica, thought maybe since they wanted to fly yours out there, they just would've offered you a sweet AMD rig in exchange for yours.

Aug, don't tell me stuff like that. I really would have loved to just ship my system off, but getting all my data moved off the machine and bringing another box up in the meantime to work on would have been a massive pain.

I'm not really in this for a free upgrade anyway. The only suggestion I've made on that front is asking them to replace my 5850 with a 6000 series card (of equal or greater performance) if the new hardware would eliminate some of these problems.

Anyway, update from AMD:
amd said:
We are in process of procuring the motherboard and the CPU. Hopefully by next week Friday we might have a preminilary update for you.
From my phone conversations with them, it sounds like they're having a difficult time getting Foxconn to supply them a motherboard to test on. I know this motherboard has a few issues of its own that were never taken care of, I wonder if they're worried AMD wont like what they see if they scrutinize a sample board... having a partner slap AMD logos all over a glitchey motherboard isn't great PR, after all.
 
Installed last night, coincided with my installation of a second 6870. No issues so far, I tried out JC2, L4D2, and TF2 without issues. Right top corner cursor freakout is still present but my second monitor works fine. I was using 11.4 beta before on a single card.

Anybody know if there's an "11.4" version of the application profiles coming out soon?

AM
 
Honestly I would respect AMD more if they fix their issues from the feedback being provided by unknown one. You'll be surprised how often people notice issues and simply don't bother sending a bug report to amd. Then the countless usueless and inaccurate bug reports lead companies to kind of lose sight of the end user feedback.

Unknown one good job bro!! + Rep
 
You'll be surprised how often people notice issues and simply don't bother sending a bug report to amd

Probably because, as you've seen, AMD doesn't care unless they're getting carpet bombed. Let's say they do care: good luck finding the correct place to submit bug reports, getting a (non-automated!) response back if you do find it, having the issue fixed, etc. You'll get more luck posting about your bug on a random forum than via official AMD channels. That's a pretty good indicator of their priorities.
 
Also, don't forget that AMD's official forums' admin (spyre) is an arrogant jerk and a class one troll who treats everyone like crap and proceeds to ban anyone who disagrees with him/proves him wrong.

So yeah, they [AMD] don't give a damn (firing him would be a good start).
 
Back
Top