AMD Needs Your Help - Driver Feedback Requested

[Warsam71]

Hello everyone,

AMD needs your help --

As we continue to improve our Catalyst Driver suite, we kindly ask you to help us by submitting any issues or bugs you have experienced with our drivers by using the form below. We value your feedback and intend to use the information you provide us to make our Catalyst Drivers better.

Note that this form, which closes on September 12, is different from the standard Issue Reporting Form we regularly use; therefore I ask that you use this form even if you already submitted to another form in the past. For the next 7 days we are making a concerted effort to increase the number of reports we receive to create a “bible” of bugs encountered by our customers.

Please click here to access the form.

Your feedback is greatly appreciated, thank you in advance!

 

[Rvenger]

Thanks Sam. I submitted a report. Having multiple issues on all my GCN cards that needs rectified. Flash player BSOD, BF4 Mantle performance issues and lots of stuttering in BF4 DX11. 14.4 is getting a little 'long in the tooth'.

 

[Warsam71]

Thank you Rvenger!

This is exactly the type of feedback we're looking for our "bible". The more the better...

 

[schmuckley]

idk but..13.9 gave me random blackouts with reference 7970.
:thumbsdown:

 

[el etro]

You need to put a Vsync switch for DirectX applications in catalyst control center. D3Doverrider is doing the job by now;

 

[Warsam71]

Thank you for your comments Schmuckley and el etro, please make sure to use form I've mentioned in my original post.

 

[Bateluer]

You need to put a Vsync switch for DirectX applications in catalyst control center. D3Doverrider is doing the job by now;

Seconded.

 

[JDG1980]

One of the longest outstanding bugs for AMD GPUs is the random colored block errors in Firefox with hardware acceleration enabled. It only happens rarely during normal operation, but if you're doing heavy GPU work at the same time (e.g. playing a 3D game in the background), it occurs much more often. Random single-color blocks appear where normal website (or browser chrome) text and graphics should be, especially during scrolling. I've found one of the easiest ways to reproduce the issue was to run cgminer.exe in the background at medium to high intensity on the same card used for the primary display. This happened on my 7750 and on two different 7870s, on multiple versions of Firefox and Catalyst. See bug 837489 which remains unfixed. One user reports seeing the same kind of alpha-blending errors in the Steam client. Firefox devs think it's the video card drivers, not their software. If you can't fix this, could you at least decisively prove who's to blame?

 

[Enigmoid]

Easy. Fix mobile drivers so that I can completely disable graphics switching in the driver control panel. This mode will completely DISABLE the igp. Enduro frankly sucks and needs a way to be completely disabled (I can disable optimus but not enduro).




It is extremely important to have working release drivers as a lot of OEM's don't allow general drivers and then fail to release updates.

 

[BrightCandle]

This is going to sound like a dumb question but if you already have bugs reported and tracked in other ways why do you want them re-reported?

I reported 22 bugs on the 7970's in January/February 2012. In that time you have fixed 1 of the issues fully. If you are genuine about fixing the bugs then that is great, but you are going to have to explain why I should dig up and re-report a load of bugs you ignored the first time and for the next 6 months when I hounded your support teams before eventually giving up all together. It was a waste of time then why is it not a waste of time now?!

 

[3DVagabond]

This is going to sound like a dumb question but if you already have bugs reported and tracked in other ways why do you want them re-reported?

I reported 22 bugs on the 7970's in January/February 2012. In that time you have fixed 1 of the issues fully. If you are genuine about fixing the bugs then that is great, but you are going to have to explain why I should dig up and re-report a load of bugs you ignored the first time and for the next 6 months when I hounded your support teams before eventually giving up all together. It was a waste of time then why is it not a waste of time now?!

Bugs from 2012 wouldn't be relevant to the questionnaire. They are looking for issues with current drivers.

 

[stahlhart]

Bugs from 2012 wouldn't be relevant to the questionnaire. They are looking for issues with current drivers.

How about you let Sam answer the questions here?
-- stahlhart

 

[DaveSimmons]

One of the longest outstanding bugs for AMD GPUs is the random colored block errors in Firefox with hardware acceleration enabled. It only happens rarely during normal operation, but if you're doing heavy GPU work at the same time (e.g. playing a 3D game in the background), it occurs much more often. Random single-color blocks appear where normal website (or browser chrome) text and graphics should be, especially during scrolling. I've found one of the easiest ways to reproduce the issue was to run cgminer.exe in the background at medium to high intensity on the same card used for the primary display. This happened on my 7750 and on two different 7870s, on multiple versions of Firefox and Catalyst. See bug 837489 which remains unfixed. One user reports seeing the same kind of alpha-blending errors in the Steam client. Firefox devs think it's the video card drivers, not their software. If you can't fix this, could you at least decisively prove who's to blame?

I get this in Firefox with my 6670 at work (latest Firefox, 14.4 cat / 14.10.x drivers) but for me it's if I allow Flash to run in Firefox. That also causes "driver has restarted" messages. Disabling Flash fixes it.

Catalyst Design flaw: why doesn't it let me copy-paste the Information tab info (manually or with a copy-to-clipboard button)? If I wanted to list the exact Driver Packaging version I'd have to hand-type an eleventy-digit code. Grrr.

 

[JBT]

14.7 RC3 has caused some corruption in Lord of the Rings: Online for me. Filled out the form.

 

[tollingalong]

idk but..13.9 gave me random blackouts with reference 7970.
:thumbsdown:

Similar problem with 14.4 and 14.7 R3; they caused complete blackouts on my 290 to the point that I returned it. I loved the card when it worked with older drivers though .

Submitted report.

 

[Elixer]

This is going to sound like a dumb question but if you already have bugs reported and tracked in other ways why do you want them re-reported?

It sure sounds like they are trying to weed out the "old" reports, and get a 'more current' list that people still have issues with.

On one hand, this is a easy way to weed out the "old" reports, but, this also means that if it don't get reported, they will assume it is fixed, when in reality it is not.

It would be MUCH better if AMD (and nvidia and...) switch to an actual bug tracker, like bugzilla or something like that, where people can look up the same issues, and then add their info.
They can still classify bugs 'private' (as in, not public viewable) if need be as well.

The current forms are only a one way street, and we NEVER hear back, no matter how many times we submit reports.

 

[HurleyBird]

You guys need to create more esoteric test rigs as eyefinity is riddled with bugs as soon as you try to do anything interesting with it.

For example, Catalyst almost always BSODs on wakeup with a 3x1 eyefinity layout + DK2, with an HDMI DAC connected to Intel processor graphics from a 4770K (trying to get HDMI receiver/DAC output out of my 7970 at the same time as eyefinity is an exercise in complete and utter frustration). When it doesn't BSOD you get corruption on the side monitors and need to reboot anyway.

Eyefinity in general needs to be better at enabling/disabling/re-enabling in more complex situations. If you try to do anything complicated you're going to need to rebuild the array from scratch every time, which is a major pain in the ass.

In general your eyefinity testing and support seems really poor outside of laboratory sterile conditions. As soon as you try to do anything unique with multiple monitors something breaks. If you want to appeal to enthusiast and pro users, this is an area you should be nailing. The only reason AMD's poor multimonitor support hasn't caused me to abandon Radeon is that Nvidia's is likely as bad or worse -- better the devil you know than the one you don't.

This is going to sound like a dumb question but if you already have bugs reported and tracked in other ways why do you want them re-reported?

I reported 22 bugs on the 7970's in January/February 2012. In that time you have fixed 1 of the issues fully. If you are genuine about fixing the bugs then that is great, but you are going to have to explain why I should dig up and re-report a load of bugs you ignored the first time and for the next 6 months when I hounded your support teams before eventually giving up all together. It was a waste of time then why is it not a waste of time now?!

I second all of this.

 

[Sohaltang]

Support for Manually setting the res and custom refresh rate.. Like um how some other company once did

 

[3DVagabond]

It sure sounds like they are trying to weed out the "old" reports, and get a 'more current' list that people still have issues with.

On one hand, this is a easy way to weed out the "old" reports, but, this also means that if it don't get reported, they will assume it is fixed, when in reality it is not.

It would be MUCH better if AMD (and nvidia and...) switch to an actual bug tracker, like bugzilla or something like that, where people can look up the same issues, and then add their info.
They can still classify bugs 'private' (as in, not public viewable) if need be as well.

The current forms are only a one way street, and we NEVER hear back, no matter how many times we submit reports.

It's been a while ago, but I've only had one bug report I submitted and I did get a response.

 

[UaVaj]

win8.1 + catalyst 14.4

three GV-R928XOC-3GD on three identical computers (i5-3570k + Z77 Ext4).
one GV-R928XOC-3GD on one computer (i5-760 + P55M-UD2).
four R9290X-4GD5 on one computer (i7-3770k + P8Z77 WS).

all having - green flash video playback which requires reboot to temporary fix.
also having - driver stop responding and recover.

not a deal breaker, but definitely buggy.

 

[Shmee]

I am not sure if it is a mantle or a BF4 issue, but I get horrible memory leaks in Mantle BF4 crossfire, so I have to use DX11 now.

 

[UaVaj]

I am not sure if it is a mantle or a BF4 issue, but I get horrible memory leaks in Mantle BF4 crossfire, so I have to use DX11 now.

x2. mantle is broken. after a game or two. needs to restart bf4.

 

[yannigr]

Guys put more people on your site.

I had to write you in the past 3-4 times to fix the page that was stating that only 260X was supporting trueaudio, and not 260. You fix it like 6 months later.

It took you only 1+ years to add 760K in the desktop comparison chart.

You finally updated the AMD Athlon's page. How about updating also the comparison page between the 2006 AMD model and the 2000 Pentium 4 model?



And about Catalyst and Sempron 3850. In your download page I was hitting my head on the wall because the results where giving me only Windows 7 and Windows 8.1 versions of the driver. But on the ASUS's motherboard page there is a driver for XP. So, if there is a driver that works on XP, why isn't on AMD's page.

 

[spaceman]

not filling out that form,
persistent not coming out of standby mode(but occassional)
id say about 1/6 times resuming machine i have to hard reset.
extremely frustrating
same issue across multiple drivers
7870myst (card was brand new as of july 2014)
other cat issues ive had:
why do i always have to mess with pixel ratios/underscan just to match my native desktop(1080p)(this is across 3-4 gen of amd cards,a more drivers revs than i can remember)
the raptr service installs too ninjaly, i hate the software.just makes the whole experience seems cheap.

 

[NTMBK]

not filling out that form,
persistent not coming out of standby mode(but occassional)
id say about 1/6 times resuming machine i have to hard reset.
extremely frustrating
same issue across multiple drivers
7870myst (card was brand new as of july 2014)
other cat issues ive had:
why do i always have to mess with pixel ratios/underscan just to match my native desktop(1080p)(this is across 3-4 gen of amd cards,a more drivers revs than i can remember)
the raptr service installs too ninjaly, i hate the software.just makes the whole experience seems cheap.

If you don't fill out the form, don't be surprised when you get ignored.