M17x-r3 (amd7970m) strange issue loosing display

I have an issue with my Alienware and would really appreciate any ideas from this community as to what to do next.

I have the model with some intel integrated gpu as well as the discrete amd 7970m. The issue I am describing only happens when the game is set to use the 7970m or the laptop is booted into 7970m only mode.

At first, the laptop would randomly crash with the screens (laptop screen with 1 external monitor) would go black, sound would be some horrendous noise and I would have to hold the power button to reboot. Upon reboot, windows would ask if I wanted to start windows normally or not, suggesting that the system crashed (imo).

Dell switched the GPU and things all seemed really good. In fact, I suspected that the original GPU was never fully functional as I was able to put games at even higher settings with the new card and everything seemed perfect.

Now, less than 2 months after the Dell service installed a new 7970m a very similar issue is occurring. But now, both screens go black but the sound of the game continues to play, suggesting that windows is still up, the game is still up, but I have lost any output to the displays. Upon a forced reboot, windows does not ask if I want to boot into safe mode either.

So I am baffled. Is this the same issue as before? While I know I should just contact Dell, I would really like to try to figure this out. But my google searches are missing the right keywords or I am alone with this issue, but I cannot find anything on the net to help.

And this has been tested both the drivers from Dell as well as latest drivers from AMD. I ran GPU-Z last night and the card was only at 63C when it "crashed/went-black-screen". My suspicion that maybe Dell did not re-paste well seems to be unfounded...

 

I am no expert on the m17xr3 but certainly sounds like hardware especially if your tried the different drivers and the temps look normal.
Can you elaborate on what horrendous noise means?

 

The horrendous noise no longer happens. That was when the laptop crashed with the previous 7970m inside. It sounded like static and noise like the computer was in a blender, I dunno how to describe it. But in any event, it no longer does that, with the present 7970m, the screens go black (actually the laptop screen is dead black, the external monitor is a series of vertical bars alternating in dark grey and less dark grey pattern). But the sound of the game continues on, it no longer makes the horrid sound.

 

Definitely sounds like a hardware issue. If you are comfortable with opening unit and warranty has expired see if any connector is loose.
Otherwise if not and still under warranty arrange for repairs with Dell.

Me thinks the fault from original repair/warranty was never correctly repaired and problem has continued.
Good luck.

 

Yeah these are typical problems caused by a failing GPU. Dell should try a MB along with a GPU to eliminate a root-cause issue. I have heard of (and experienced) multiple GPU failures but that was with the NV 580m/675m which is a known poor GPU .

Good luck with Dell and please post results .

 

So the theory here would be that the mobo is faulty and can kill a brand new 7970m in 2 months with minimal usage (I play a little at night and mainly on the weekends)?

I cannot dispute that, but the strange thing is the randomness of the black-screen/non-crash events. Last night, with GPU-Z running I was trying like mad to get it to crash in Grim Dawn by pumping the settings to maximum. Temps stayed 60-63C and it would not crash, I went into a match in WoT where I had the game still in the lowest settings as I was previously playing it on the integrated GPU, and (with the AMD card) it crashed right away, temps still 63C max. As long as I do not use the 7970m using configurable graphics, it never crashes. When the AMD is active, a crash is not guaranteed, but I would peg it about 90% chance.

In any case, thanks for the replies. I will definitely keep this thread updated as the story unfolds.

 

Yeah re the mobo. Just a guess but one dell have done in the past. The randomness is nothing new with a failing GPU. A game just has to 'hit' the failing code once too often and it flakes out.

With me after two GPU's and a mobo they offered me a new machine and I got the R4 in my sig (from an R3). Unfortunately the GPU was still a weak point and I had two more failures before I replaced it with a 680m that has been rock solid.

Good luck

 

Well, no idea how to change a thread title, but it turns out I have an m17x R4. Dell told me to upgrade BIOS. I had A10 and have just put on A12. Going to see if it still fails now, but am curious if anyone here expects a BIOS update to be the fix?

 

Probably not, sounds like a dying card.

 

I agree, not much chance but the first thing dell want with a GPU problem is their driver and the latest BIOS. Then they can proceed.

 

Well, about 3 hours playing games that normally crash it and still nothing, if this BIOS update really did it, I will be satisfied, Grim Dawn is still totally playable with all settings maxed to the max.

Thanks for the replies guys. Will update here if it crashes again and we call Dell for the part switch operation.

 

No dice. I had to reinstall diablo3 to goto westmarch heights, the place I first encountered this crash, and it still crashes. Dell will be visiting soon.

 

Ok, to be comprehensive I update with today's news. This is more of a rant than anything else, but for others with Dell-experience, I am sure it is fun to read about it happening to someone else.

From ~9:30am to 11:30 am I was on the phone with Dell who I let access my m17x remotely. I had to sit there on the phone with him and basically just enter my password each time he rebooted. He found the crash logs from last night, proceeded to download the A12 bios (which in chrome showed the classic (1) at the end of the download file name cause it was already there. I stopped him on that one and then he did the same thing with the latest AMD driver...I stopped him again. I did not stop him (cause I was at another computer) in time to run the custom installer, so he installed all the crapware from AMD on my machine in addition to the driver. Re-ran some lame tests that did not even get the fans spinning, booted into f12 and ran more lengthy tests and concluded there is nothing wrong. But he saw the crash report and thus does not deny there is an issue, just that he cannot figure out what it is.

I am to wait until the next crash (planned for tonight when I can re-run that d3 map) and then they will schedule a meeting for the guy to come with parts...

 

What country are you in? This is typical Dell support and not what their AW division used to supply. The next thing will be a re-install of the OS and connection to a clean machine to verify only dell drivers (even if you have done this). Then you game again and the problem is still there. The next step will be them wanting you to send it in. They cannot identify what part needs replacing so it's all in the warranty .

I say this as worst case and really hope they just send a GPU and mobo as they have done for people with similar problems in the past. I had a cable problem to the display. The 'alienware' logo did not light. Now it could be the cable or the panel unit led. They sent both the cable and a full panel assembly. The engineer fitted them both - 'while I'm there'. They could have just said 'send it in' but went the extra mile which made me very happy with the support operation .

Good luck and thanks for the update .

 

This is in France and probably even a special subdivision of Dell here for academic research labs. When they previously changed the GPU, they sent a guy to my office with the part to change it. Funny thing was the first part was obviously used and did not work. He returned again the next day with another box of GPU and this one the bracket was not already attached and works 1000x better than the original one ever did.

So I expect they will send the guy with the parts to my office (at least for this next visit, if it is not the mobo/gpu pair, then all bets are off what is next). I just wish they would send him with more than 1 possibly refurbished part this time so if the replacement is doa, we can try another.

 

Well, I have run this particular map in d3 three times now without the crash-incident. But in doing so, I notice something bizarre. I cannot remember for sure if GPU-Z was showing fan speed before but it sure is not now and I really think I would remember things like "-%" and "-RPM" showing in the display. The reason I started GPU-Z is cause the laptop seems too quiet now. In my office earlier when Dell was connected and running their test with spaceships flying around in space I also noticed that the fans were not kicking in. There I figured it was because the test was not really that taxing. But in d3 just now I had GPU load at 99% and temp at 66C but the Fan speed readings stay at "-". A fan does kick in but I cannot be sure which one and the GPU goes back to 62C or so quite rapidly. CPU temps maxed at 64C or 67C depending on the Core (loads between 35% and 50%).

Anyway, really not sure what is going on at the moment and would appreciate any pointers on what I should be doing to try to diagnose this.

I did not follow every step the Dell rep did while in my computer today. I know that he used the AMD download to uninstall the previous driver, removed the device after reboot and then installed the driver after another reboot. I do not think I ever did those steps exactly like that, but I do remember attempting a method involving other tools, something like DDU or something that completely removed tons of stuff from my computer including steam login info and the like, not just AMD entries in the registry or whatever I thought it was supposed to do.

This problem has been going on for quite some time and so many things have been tried I almost feel like I am loosing my mind with respect to the details over the entire time course. For now, I just want to make sure the fans are working cause maybe the A10-A12 BIOS upgrade changed something? Or maybe the way the drivers are installed now has possibly fixed my problem but also caused GPU-Z to no longer monitor the GPU fans?

 

Definitely looks like a hardware fault on the motherboard especially if the fans are not kicking in that sounds like sensors have fried and it cannot tell the GPU fans what speed to run at.
Assuming the motherboard bios is the latest on your system and everything else is normal nothing unusual then the motherboard will need to be replaced with Dell.

Continue to request repair with your local Dell support. Also, identify the fault from the last time has not been repaired and it continues to get worst. Also ask the warranty to be escalated or talk with senior supervisor at Dell.

Good luck.

 

Well, after a decent bout of testing, I have not had a crash incident. I think I might know the solution in my case. Sadly, I am not 100% sure of the dependency of the BIOS upgrade part of it.

If anyone else encounters black screen crashes with the 7970m, I would suggest, with or without BIOS upgrade if you want to test, the following:

1) Upgrade BIOS to A12 (I was on A10).
2) Use AMD.exe to uninstall CCC/driver (just the default Uninstaller).
3) Reboot and use device manager to remove the device
4) Reboot and re-install AMD.exe (Dell rep used the latest Catalyst mobile driver from AMD site [amd-catalyst-omega-14.12-with-dotnet45-win7-64bit]).

Since the m17x arrived, I think mine has been badly installed. I have no clue if the BIOS played a role or not. I did that first before I imagined it was not a hardware fault.
But I have installed drivers on this thing many times in many ways following guides on this forum to be as clean as possible. So it might be that the combination of the BIOS upgrade and a clean install of the latest AMD drivers on top of it, finally got my GPU working properly. Temps stay cool, max 66C and fans are working, the do not start when the laptop is cold. But after gaming an hour, they start to kick in from time to time. GPU-Z is still not reading them, but they work and the laptop seems to be working well.

I will update as soon as the next crash happens if it does. I have Dell on standby for an intervention, but for the time being, no crash.

 

What do oyu mean by this? Is the laptop crashing due to high temperature after an hour?

 

It was never a problem of heat. But, and I know nothing of drivers and hardware and the lot, I begin to suspect that some issue between drivers and bios were amiss before, causeing fan activity before it was needed and maybe even cutting off the card because it was misinformed that there was heat issues.

the issue always seemed to be gpu shutdown cause it was black screen accompanied or not by hard crash, reboot power button only. but hardware scans fail to find error and temperatures were always in the 60s.

 

I'm experiencing the same problem (I think) and followed your advice above. Will update if anything breaks.

 

How frequently were thesse blackscreen crashes occuring? And did you mean that the screen looked liek a dark green color?

 

Frequency is tough, I would say I could 100% of the time get a crash on a specific diablo 3 map, and 99% of the time playing an old Star Wars RTS, then Grim Dawn and WoT would crash about 50% of the time, if I played long enough, it would happen, but never with any rhyme or reason as far as I could tell. But I was essentially no longer playing with the discrete card cause I knew it would crash on it.

I play on an external monitor with the windows desktop extended to the laptop screen. The laptop screen would always be totally black. The external sometimes black, other times more dark-green-grey vertical bars could be visible.

Anyway, for now, my machine has not crashed once and I have been leaving games on all day long cause I have Dell waiting to send someone. But if it does not crash, I am not really tempted to initiate the next visit. I guess I sort of fear "new" problems from "new" parts and as long as the current parts work...

 

Interesting, please keep is updated with any changes please!
Currently buying a 7970M, and would love to avoid any potential problems.

 

For now, still no issue. Been playing a lot and everything seems okay. Just wish I knew why...

 

For me its looking like a really bad AMD drivers (and they suck so much that they could cause that issue) or dying GPU. You could try underclocking the GPU and check if the black screen appears.

Don't know if you know but previous Radeon HD6990M has died in every laptop. I don't now know single person that has 6990M working after 3-4 years. Mine died after 1,5 year. 7970M is a lot better than 6990M but I still REALLY recommend buying laptop with Nvidia GPU as most of poeple here switched from 6990M/7970M to 680M/780M/880M(like me)

 

I spoke too soon, within 1 hour of my post, it crashed again.

I have just sent Dell an email, saying I am done with this laptop. I will update with their reply when I get one, but I do not want to see this laptop get a new part, I want a whole new one, I have had it.

 

Well, this story is still ongoing and I hope to some time in the future update here with all the details. But the reason I am posting now is the I downloaded CCC 14.8 from Guru3d and did a clean install of that driver and have not had the issue since. This will require at least a month of testing before I can really claim that it is a fix, but so far so good.

http://www.guru3d.com/files-details/amd-catalyst-14-8-whql-(14-201-1008-august-12)-download.html