Radeon HD 7970 CrossfrieX Issues

Final8ty.

A good troll always need to justify himself.
I explained myself and it seems you just got the wrong end of the stick and i don't need to justify myself to anyone on the forum.

You have yet to tell why its a troll comment as to me it just looks like you just didn't like what i said even though its the truth and it seems you presumed i pointed it at you, so you call it a troll comment, the easiest thing to do is slap it on any comment you don't like, exacerbated by the fact that your having such CF eyefinity issues yourself .
Now if you have anything to back up your opinion that im trolling and that my statement was false and that most do use the AMD's issue reporting form and don't moan on forums then go ahead or i would take it as you being the very thing your calling me, it was certainly sarcastic but trolling far from it, where have you been if you think that is trolling because really regardless of what you think me-
Didn't know there was so many Trolls on [H]
Is fare from the truth.
 
Last edited:
Sorry if I didn't read your comment correctly. I thought you were pointing at me when saying I was bitching on forum, I am making a lot of effort to get this issue sort out. And when I said its sad that Dan-D surrender, its because this guy is a reviewer, he has far more weight than me in the community. Thats why its a shame to loose a good teammate.

I really hope this issue gets solved. (Lets not get away from the topic)

I am confident that it will be solved next week.
 
Sorry if I didn't read your comment correctly. I thought you were pointing at me when saying I was bitching on forum, I am making a lot of effort to get this issue sort out. And when I said its sad that Dan-D surrender, its because this guy is a reviewer, he has far more weight than me in the community. Thats why its a shame to loose a good teammate.

I really hope this issue gets solved. (Lets not get away from the topic)

I am confident that it will be solved next week.

Well i have been following your issues on the other forum and the effort you have been putting in, but really AMD has to get it sorted as its been long enough.
I had Quadfire 3870 now Quadfire 5970, i would be surprised if the 7xxx defeats AMD in sorting it out, but the last few drivers have seem to be pointing in the right direction with some users on another forum saying that its the first time that its working as it should in CF but there is still much to be done.

Anyway no hard feelings.
 
Last edited:
(Its suppose not to be the same drivers as the asder setpember 4th or August 15th Mod version of 12.x(9.00).

I hope not, I had nothing but issues with the both of those drivers. 12.8 has been good to me but I'm not an eyefinity user. I'm just really hope that the crap doesn't carry over from those drivers. AMD's betas and hotfixes have generally been good to me so I have hope.
 
Which would be saying that most do use the AMD,s issue reporting form which is not the case even though you and I do personally, so no im not trolling , im just stating the facts because there has been a bigger push to get people to use it as it was not getting used enough.


But that was ok and not trolling ? personally don't think it was and mine was not any more so.

Just one without subsequent follow ups is usually taken as not that big of an issue or resolved.
The quadfire eyefinity issues is getting addressed because of the people who constantly use the AMD Issue Reporting Form and if they are airing on a forum they are on the right forum and [H] is not it.

So far the issue remains unresolved or even really acknowledged by AMD. I've submitted bug reports like this and I've gone higher up the chain in the past with issues and I've been ignored. Take that for what you will. I'm just stating that given my experiences, I don't expect an adequate response from them. Ever. They still won't do anything about the ridiculous cursor corruption and RDP issues that have plagued the drivers since the Radeon HD 5970 days if not longer.

Hey Dan_D, I'm running a Gigabyte x58-UD5 with two gigabyte 7970's stock speeds, 12.8 cat drivers. I'm having the same issue you seem to be having. I have tried different settings but continue to have the same issue. Did you ever get yours fixed?

The resolution for me was ultimately replacing my AMD Radeon HD 7970's with NVIDIA GeForce GTX 680 4GB cards. I was tired of not being able to play games and I just wamted my computer working again.
 
Which would be saying that most do use the AMD,s issue reporting form which is not the case even though you and I do personally, so no im not trolling , im just stating the facts because there has been a bigger push to get people to use it as it was not getting used enough.

What issue reporting form? It just says this now.

This survey is now closed. Thank you for your interest.

Stuff like that irks me when AMD doesn't even have a forum to report issues at like Nvidia. Having a representative to talk to the public really helps your companies image at least in the eyes of customers and thats one place where I feel that amd really drops the ball.
 
So far the issue remains unresolved or even really acknowledged by AMD. I've submitted bug reports like this and I've gone higher up the chain in the past with issues and I've been ignored. Take that for what you will. I'm just stating that given my experiences, I don't expect an adequate response from them. Ever. They still won't do anything about the ridiculous cursor corruption and RDP issues that have plagued the drivers since the Radeon HD 5970 days if not longer.



The resolution for me was ultimately replacing my AMD Radeon HD 7970's with NVIDIA GeForce GTX 680 4GB cards. I was tired of not being able to play games and I just wamted my computer working again.

On Rage3D there are at least 4 official beta testers and feedback from AMD through them.
As for those 5970 issues I haven't seen them on my setup since the mid 11 CATS so they fixed it for me.

As for others having 7xxx eyefinity with CF issues or even if you don't have the issue maybe give this thread a shot and post as they need the feedback http://www.rage3d.com/board/showthread.php?t=33994670
 
Last edited:
What issue reporting form? It just says this now.



Stuff like that irks me when AMD doesn't even have a forum to report issues at like Nvidia. Having a representative to talk to the public really helps your companies image at least in the eyes of customers and thats one place where I feel that amd really drops the ball.

I have posted many times about NV forums being much better and image that it sends, i even said that in this thread already and is the NV forum back up.
 
Last edited:
Well I bought a separate NON-ASUS board, just to try it out. Same lockups on both mobo # 1 ( Asus Sabertooth X79), mobo # 2 ( MSI Big Bang Xpower II ). I've also tried my cards on my brothers P67 system - they work there just fine, and I've tried a different PSU ( my bro's)
Bascially the list of what I've done and what didn't fix the problem:
1. Not the cards - work in my bro's P67 system
2. Not the mobo - swapped my sabertooth x79 to MSI Big Bang Xpower II
3. Not the ram - swapped two sets
4. Not the PSU - swapped that **** as well
5. Not the HDDs - swapped 3 of them\
6. Not the CPU - swapped Gen 2 and Gen 3, and ran OCCT and Prime for 12h stable
7. Not Crossfire bridges - swapped 2 sets
8. Not mini-DP cables - swapped active adapters


That pretty much leaves the Drivers

Also : 12.6 12.7 12.8 12.9 betas -> lockups in Metro 2033 benchmark within 1-15 loosp. RC11 - 60 loops strong and going on the same system. RC11 also work in BF3, but fails in other games with same lockups. And lockups happen only in eyefinity + crossfireX

So yeah, fuck AMD for not fixing a presistant issue that seems to be around since release date on x79 platform ( since cards work fine on P67 )
 
Thanks for this info! When I bought my 7970 cards I didn't want to change my computer, so I installed them on my 3 years old X58 system, and the crashes started. I reformatted, reinstalled... tried everything until I gave up and sold the motherboard, CPU and memory, then bought the X79... and didn't change anything.

After all the reports I have collected on different forums, to me it's more and more clear. The crashes only happen with 79xx cards in X58 and X79 systems. My Z77 computer remains rock solid (in PCIE 3.0 by the way).

Got new info from AMD that I'm allowed to share. AMD has identified a potential compatibility issue with a specific manufacturer’s hardware and is working closely with them to resolve the issue. The issue is receiving the highest attention at AMD and they hope to have a solution very shortly. AMD wants to thank all the users that have reported this problem, their feedback has been really valuable for them.

http://www.rage3d.com/board/showthread.php?t=33994670
 
Hope this is fixed soon my 7970 are on the way and I waited for the drivers to get better.
I never buy cards when the first come out always wait for the drivers to mature but 10 months damn.
Btw is the less problems running in 3240x1900 Over 5700x1080?
And my cf 6970 ran eyefinty fine on my x79 rive.
 
Hope this is fixed soon my 7970 are on the way and I waited for the drivers to get better.
I never buy cards when the first come out always wait for the drivers to mature but 10 months damn.
Btw is the less problems running in 3240x1900 Over 5700x1080?
And my cf 6970 ran eyefinty fine on my x79 rive.

6970 old technology and optimized over 6+ years.
cant compare to 7900 series new tech arc.
it will be 1-2 years more until its all the same as the older tech.

I have no issue with my 7970 single card and 5040x1050 gaming.
one more nm shrink and it be pretty much it.
 
It was an off the cuff comment. And I detailed out what I did regarding my situation. I went farther than most would have. And the way that is worded it sounds like the fix will most likely come in the form of a driver update, or a firmware update for the graphics cards. It may also come in the form of a BIOS / UEFI update. We really don't know anything more than the fact that AMD has finally acknowledged a problem with the hardware.

We also don't know if this could have been avoided by a better QVL process on AMD's part. In any case I have to go with the problem being an AMD one for the simple fact that AMD cards don't work together in a system, or multiple configurations and simply switching to NVIDIA cards and drivers resolves the problem instantly.

So while I don't claim to know what the actual nature of the problem is, I believe I've satisfactorily isolated it to being an AMD issue. It may not be entirely their fault, but if it's their cards are the one variable that changes and is the difference between a system that isn't usable vs. one that's stable for hours and hours of gaming, I don't think laying the blame largely on them is a mistake, a lack of observational ability, or poor troubleshooting. If it were a specific motherboard manufacturer's design flaw, the problem wouldn't exist on so many different configurations. Especially when X58 and X79 seem to be the common theme where problems are concerned. Seems Z77 and other LGA1155 chipsets aren't effected or aren't as likely to suffer from the problem.

And I'm not walking on egg shells about it. Somewhere in the design of those cards, be it drivers, firmware, hardware, or whatever there is something not right. NVIDIA's made plenty of mistakes too, so I'm not playing favorites or anything. I've come down on them pretty hard about virtually any card with "GX2" in their naming convention. The GTX 590 was a piece of shit, etc. But AMD fucked up something with the 7970's where CrossfireX is concerned. At least that's the conclusion I've reached until I get more information which alters my opinion on the matter.


The point is it was said to you before that there is a possibility of motherboard compatibility issues but you would not accept that at the time.
From what i have seen over there years, 7xxx issues was far from the norm with some setups flat out having issues no mater how many drivers it had gone through, i made the collision a long time ago that there is a serious compatibility issue and not that it must be just shoddy coding on the drivers and that there is a deeper problem.

Lets hope it a fix can be put into the drivers once they know what it is exactly as that will reach more people quicker than a mobo or even gfx card bios update..
 
The point is it was said to you before that there is a possibility of motherboard compatibility issues but you would not accept that at the time.
From what i have seen over there years, 7xxx issues was far from the norm with some setups flat out having issues no mater how many drivers it had gone through, i made the collision a long time ago that there is a serious compatibility issue and not that it must be just shoddy coding on the drivers and that there is a deeper problem.

Lets hope it a fix can be put into the drivers once they know what it is exactly as that will reach more people quicker than a mobo or even gfx card bios update..

Quoted for the comprehension failure. There was that guy crying: "It's an ASUS problem" over and over again that I disagreed with that. I even tried an EVGA X58 board and that didn't resolve the issue. I never said it wasn't a hardware compatibility issue. I only stated that if there was, it wasn't an ASUS specific issue. And it doesn't sound like it is. And again we do not know the form the fix will take. Yes it seems AMD has had to work with motherboard manufacturers to resolve it, but for all we know they just have to code the drivers differently to resolve the issue. Or it may come down to a combination of BIOS and driver fixes to resolve the problem.

And as far as compatibility goes, I acknowledged the fact that this was a real possibility. Though I believed that it was more likely a video card compability problem with the fault resting primarily on AMD's shoulders. Not the shoulders of the motherboard vendors. Especially when ASUS boards and systems are not the only ones effected, and some ASUS based setups work just fine. In fact currently it seems to be more of an issue with X79 and X58 based systems than anything. Z77 based setups seem to be working fine.
 
And as far as compatibility goes, I acknowledged the fact that this was a real possibility. Though I believed that it was more likely a video card compatibility problem with the fault resting primarily on AMD's shoulders. Not the shoulders of the motherboard vendors. Especially when ASUS boards and systems are not the only ones effected, and some ASUS based setups work just fine. In fact currently it seems to be more of an issue with X79 and X58 based systems than anything. Z77 based setups seem to be working fine.

As i said we shall see, as the point was about compatibility which is not the same as a few clumsy coding errors in a driver, personally i would not put the blame on anyone when i don't know what the specific issue is and why as there is no facts to base blame on.
Round hole square peg, which is at fault who knows.
 
Last edited:
As i said we shall see, as the point was about compatibility which is not the same as a few clumsy coding errors in a driver, personally i would not put the blame on anyone when i don't know what the specific issue is and why as there is no facts to base blame on.
Round hole square peg, which is at fault who knows.

Fair enough, to some extent but this seems like something that should have been caught during the QVL process.
 
Fair enough, to some extent but this seems like something that should have been caught during the QVL process.

Which would assume that the relevant mobos were being used during the QVL process and the issues did not present itself.
Its not like its triggered by a rare combination of unusual user events, its just everyday tasks and gaming that makes it fall over so its very unlikely that they had the effected mobos.
 
Last edited:
Which would assume that the relevant mobos were being used during the QVL process and the issues did not present itself.
Its not like its triggered by a rare combination of unusual user events, its just everyday tasks and gaming that makes it fall over so its very unlikely that they had the effected mobos.

In the end we just don't know anything about the QVL process used by some manufacturers. I have some idea about ASUS processes for their motherboards, and for companies like Tyan and Supermicro, but I couldn't tell you anything about how AMD or NVIDIA do it.
 
Well I'm having the same CFx lockup issue... I have a trifire setup. Can anyone suggest a good driver version that works? Right now when I play GW2 im getting on average 50% usage on my 3 cards, if i switch to just 2 card Crossfire, i get 99% and it's so much smoother. I figure it's driver related, and the lockups as well, but I'm at a loss as to which drivers to use. I've spent well over 4k on this system and can't play a game for more than 10 mins without having to do a long power button press to shutdown and restart the system.
 
Well I'm having the same CFx lockup issue... I have a trifire setup. Can anyone suggest a good driver version that works? Right now when I play GW2 im getting on average 50% usage on my 3 cards, if i switch to just 2 card Crossfire, i get 99% and it's so much smoother. I figure it's driver related, and the lockups as well, but I'm at a loss as to which drivers to use. I've spent well over 4k on this system and can't play a game for more than 10 mins without having to do a long power button press to shutdown and restart the system.

It may not be a driver issue. Go back and read through the thread if you haven't. It very well could be a hardware compatibility problem. I couldn't make those cards work in my system for example. It didn't matter how many drivers I tried and I had tried 12.1-12.8 without success. In fact the drivers rarely changed anything in regard to how long the system could operate before locking up.
 
I have read them, it's hard to say what it is, has anyone got these lockups in windows? while just working in windows environment? I have. after 12.9 came out I installed them over my previous drivers and in Windows 7 x64 ... I was getting Explorer.exe Force closing because of one of amd x64.dlls (cant remember the exact name of the dll) but it was obvious its from amd. I then switched a couple days ago to Windows 8 to see if I get the same problems...

Also with the hard lockups, can it be a direct x issue? because after windows 8 granted i didnt really play much GW2 but i didnt get any lockups but I did get bad scaling so i installed msi afterburner to see my gpu usage which then installed all directx versions from the last 10 years and as soon as i got into my game not even a couple mins passed and it gave me my first hard lock. Just throwing this out there...
 
I might just reinstall windows 8 tonight and use 2 card cfx not trifire (since i get 100% usage on 2 cards) and see if I get the lockup, because before I installed msi afterburner I had only played for about 20 mins one night and another 20-30 mins the night before which i sitll consider a good gaming session when it doesn't lock up ;) but as soon as afterburner with directx was installed it locked up.

Anyone think its worth reinstalling? Or anyone already tried this?
 
I have read them, it's hard to say what it is, has anyone got these lockups in windows? while just working in windows environment? I have. after 12.9 came out I installed them over my previous drivers and in Windows 7 x64 ... I was getting Explorer.exe Force closing because of one of amd x64.dlls (cant remember the exact name of the dll) but it was obvious its from amd. I then switched a couple days ago to Windows 8 to see if I get the same problems...

Also with the hard lockups, can it be a direct x issue? because after windows 8 granted i didnt really play much GW2 but i didnt get any lockups but I did get bad scaling so i installed msi afterburner to see my gpu usage which then installed all directx versions from the last 10 years and as soon as i got into my game not even a couple mins passed and it gave me my first hard lock. Just throwing this out there...

I had no trouble in the Windows environment. Only in games. It's most likely an X58/X79 motherboard compatibility issue. Those two platforms seem to have the most trouble with CrossfireX by far. Z77 may have issues too, but I don't know of any Z77 Express based systems having this kind of trouble.
 
Can the problem be firmware related? What firmware does your bios use?

If it is, there is no fix for it currently. I tried every BIOS available for the reference cards I had, as well as my motherboards. I had the problem on three different boards, from two manufacturers using two different chipsets.
 
I've been keeping an eye on this thread and I've noticed that both you and yurdun have 32 GB RAM. Did either of you try reducing the amount of memory to 4GB?
 
I've been keeping an eye on this thread and I've noticed that both you and yurdun have 32 GB RAM. Did either of you try reducing the amount of memory to 4GB?

That wouldn't have any impact on the problem and makes no sense. Even so by virtue of testing on different platforms I've tried 12GB and 8GB of RAM and it didn't change the outcome. I still got hard locks around 30-45 minutes or so. Sometimes faster.
 
That wouldn't have any impact on the problem and makes no sense.

What if something were having a problem operating beyond 4 GB? An addressing issue, perhaps, or a faulty trace inside a chipset.
 
That wouldn't have any impact on the problem and makes no sense. Even so by virtue of testing on different platforms I've tried 12GB and 8GB of RAM and it didn't change the outcome. I still got hard locks around 30-45 minutes or so. Sometimes faster.
What if something were having a problem operating beyond 4 GB? An addressing issue, perhaps, or a faulty trace inside a chipset.


There was a more than 4GB bug that hit AMD once that would stop the drivers installing on more than the first GPU in multi GPU setups, and that was when 8GB was the exception, but they did fix that on the next driver.
 
i disabled trifire and just did crossfire and I have not had a single hang in over 4 hours of straight play or actually all day of playing around 7 hours or so...when I put it on trifire it died within 20 mins or so. I have to debug further, going to try using the other card in cfx to see if its my 3rd card that is making it hang or if its the combination of all 3 being used at once.
 
i disabled trifire and just did crossfire and I have not had a single hang in over 4 hours of straight play or actually all day of playing around 7 hours or so...when I put it on trifire it died within 20 mins or so. I have to debug further, going to try using the other card in cfx to see if its my 3rd card that is making it hang or if its the combination of all 3 being used at once.

It would be good to know. I wonder if it might be a specific hardware revision of specific cards that are at fault when used in combination with certain motherboards. AMD probably won't disclose much about the issue. Ever. So I'm really curious about this.
 
I just picked up a 7970 in a trade, it's sitting in the corner but I'm eventually going to throw it in and compare to my 670...and wanted to try 7970 xfire...this is discouraging lol.
 
well I tested my other card as well using both of the cfx adapters that I was using for trifire. both cards and both adapters played fine as long as I was in cfx not trifire. I played over 2 hours on each configuration. it seems for me only trifire fails, maybe different drivers for me will fix this?
 
I've been keeping an eye on this thread and I've noticed that both you and yurdun have 32 GB RAM. Did either of you try reducing the amount of memory to 4GB?

well, if it is any help i've had hangs on 2 different systems with:
4 gigs ram
6 gigs ram
12 gigs ram
16 gigs ram

so i don't think it was the ram.

ps: after switching from firefox to google chrome i can finally say i've been 100% crashfree ever since (about a month now).
Was one hell of a battle though, took me 9 months and alot of frustration to get the system this stable.
 
nevermind... I got back on to play and I got a hard freeze with my third card 2nd cfx adapter... gonna go back to second card 1st adapter to see if I get the freeze... gonna play like that for a week to see if anything... maybe I just got a lucky 4 hours of problem free gaming last time, ill keep updating if any breakthroughs happen
 
I've only encountered one problem with my system in the 12.9 beta drivers.

My explorer.exe crashed because i was a different resolution when i was playing the game vs on windows (bezel compensation on vs off). this caused explorer to crash, but it hasn't happened recently.

Thinking of doing a re-install of windows, but i might just do a ghost of my drive before i wipe it.
 
Well after about a week of playing on my first configuration and trying both of my cfx adapters playing in crossfire not trifire, everything has worked without a problem. This leads me to believe that its 1 of 2 things at least on my system. The third card is bad or the pci-e slot is bad. unfortunately im running on water cooling which will make it difficult for me to debug this any further unless I remove and redo my water setup and swap 2 of my cards to test the other card in the other slot. I'm more inclined to just RMA the card saying it's faulty under any cfx or trifire setup. I would think that if it were the board my other 2 cards would also give me full freezes while playing but it seems as long as I remove my third card from the mix everything runs smooth.
 
Back
Top