I hope someone here will have suggestions on how to fix this problem.
I have an M17x R3 with 580M and a premium LCD screen for 3D. The warrantee ran out about a year ago and just before then, the LCD screen was replaced along with some repairs to the frame. At that time all the drivers on the system were also checked and any needed updates made. Since then the only driver I have updated is the video.
From that point, the computer ran flawlessly until 11/11/2013. It then had a single crash then worked normally for several days. Since then the crashes have become more and more frequent. Initially I thought they might be heat related and it is possible the computer overheated at some point, although I had used compressed air to blow out the vents I did not open it until the crashes became more frequent and the heatsinks had accumulated lint.
My initial diagnosis based on the lint was a heat-related video card failure. I cleaned out the lint with the hope that maybe nothing was damaged and it would work ok once the overheating was eliminated.
However, the computer continues to run off and on, sometime experiencing periods where it will run fine for a full day, other days crashing within minutes everytime I try to boot it up or even going to the 8-beep error code. It seems to crash faster when there is a heavier load on the system (crashes sooner if attempting to play Skyrim than it does idle or browsing).
Hardware testing during longer stable periods has come back with everything checking out fine. Running a GPU status logger records no temperatures much over 60C while online sources suggest it should run to 77C before the overheating limiter kicks in. No significant changes in the GPU status (temp, clock speeds, memory usage) occurred before crashes. Given how far below the limit this is, I'm not sure the amount of lint I cleared out would have caused it to get to the point of causing heat related damage.
Most of the time it runs stable in safe mode -- I have yet to experience a crash while running in safe mode, but have not left it running in that mode for as long a period of time.
I have been using NVidia's GeForce experience to stay on the latest drivers. I noticed there was a new driver around the time the crashes started so tried doing a roll-back. For a couple hours I thought that had worked then the crashes started again. It was at this point I first experienced the 8-beep code.
An 8-beep error code is indicated to be an LCD failure. I'm not sure how an LCD failure would cause the entire system to shut down rather than the screen just going blank. I'm also not sure whether it might simply be a loose connection on the LCD since it seems intermittent.
Anyone have suggestions on what to try at this point? If it's the video card I'll replace it, but I'd hate to replace the video card then need to replace the LCD or vice versa, or replace them and still not be able to fix the problem, or replace them and find out it's fixable without a major hardware purchase.
Update: It seems to work just fine in safe mode as it ran for hours with no problem yesterday. I'm not sure if that means it is a driver/software problem rather than a hardware problem or whether it could simply be that safe mode also keeps it away from triggering a hardware problem in some cases.
Before spending any money on a replacement GPU, I would disassemble it, remove the graphics card thermal assembly, clean the thermal paste off of it and the GPU. Apply new paste. Test it. Use the online Service Manual and watch this video:
Right now the GPU isn't coming anywhere close to overheating. Although I initially expected an overheating problem, the cooling is keeping the GPU at a consistent 60C or lower. Are you suggesting that parts of the graphics card could be overheating even if the GPU is running cool? I can see that could be a possibility.
Running in safemode ignores the 580m and is just using the onboard Intel HD3000 with a VGA driver. That also leads me to think that the culprit is the 580m or the paste itself.
That explains why the 580M could be the problem that is avoided when running in safe mode.
I've been using GPU-Z to log the temperatures on my tests out of safe mode. The crashes happen there regularly but the tracked/logged GPU temp never gets over 61C and only touches that for short periods of time. This is puzzling to me as the GPU should actually be running hotter than that based on the specs I've seen.
When it isn't crashing, it also passes all the video card tests included in Alien Autopsy.
I'll look at getting some paste and trying to remove/clean/paste/reinstall the video card.
I have the same laptop as you, and my issues started happening around the same time..i.e. sometime in November.
I've been playing FFXIV: ARR since beta testing in July through November when all of a sudden my screen went black, then rebooted. I thought it was a fluke thing because it didn't happen again for a couple of days, then it started becoming a regular thing. Sometimes it would happen after playing for a couple of hours, sometimes before I even left the login screen.
At first I thought it was my hard drive. A couple of months before this started happening I put a SSD in and it was a little quirky. It would BSOD upon waking from sleep, and every fifth boot it would get stuck in the loading Windows screen. I called Alienware, even though it's been out of warranty for about a year, and they had me run AlienAutopsy, which I didn't know about. The HD came back as faulty, everything else passed, so I put the old HD in and reinstalled Windows/drivers/updates. My computer continued to crash, only in my game though.
All my temps are fine. GPU in the low 60s and CPU in the low 70s. I also ran memtest and AlienAutopsy again and everything came back okay. I played GW2 for a few hours and it didn't crash so I just assumed it was FFXIV, until I was checking my email a week ago and it happened while in Outlook. It was then that I decided to take it to a repair shop. Well, they ran all sorts of tests, stress tests, and played games on it without it crashing. Their diagnostic tools suggested that the video driver was causing the crashing, so they updated the video drivers. I thought this was strange because I was certain that I had already done that, although nvidia seems to release new drivers fairly regularly, so who knows. I went to the shop to pick up my laptop today and they had me play my game for a bit before I left. It played fine. I paid my thirty dollars, their fee, brought it home and played for a couple of hours with no issues. Then later in the evening I played for ten seconds and it crashed. I'm at a total loss. Tomorrow I will be contacting the shop again, but at this point I'm not sure what can be done, or rather, what they can do. I did run their "What Crashed" (not sure if that's the exact name) program they left on my computer and it was an nvidia 331 driver kernel responsible for the crash.
I will let you know what they say/do, and please do the same!
Still having trouble with this. I took it to a local shop that determined the video card (580m) was the problem using a basic stock card they had there to test it. It seems like there are a lot of problems with the 580m's and Dell is currently out of stock...based on posts here they were also out of stock a few months back.
The shop thought a GTX 760m would work in my computer and in a similar Alienware M17x-R3 with a 560m that came in at the same time mine did. Although x60m is a big downgrade from x80m with 2 generations newer it would be a significant net upgrade so that sounded good to me. The 760m also uses less power and should heat up less than the 580m.
I understand from looking around the forums that people have done similar upgrades, but the shop has not been able to get the cards to work in either computer. now. Any recommendations?
Apparently GTX 580m's are tough to find now and installing anything more recent requires a BIOS reload/unlock. The shop considers that too risky to make sense for them to do, so they have punted it back to me.
I'm back to planning to do the work myself and I'm researching the unlock process as well as video card options. I'd love to throw a 780m in there, but that may be more than I'm willing to pay right now and I'm concerned about the power load. 760m and 770m both sound like they might be good options.
Any recommendations? Maybe this should be a whole new topic.
After reading your post due to experiencing the same issues. I bought by R3 approx. 2 years ago and since then I have had nothing but issues with it. I have had replacement heat sinks, gpu, cpu, keyboard and now since around November I am experiencing the same issues your are mentioning.
I tend to have issues in games and the laptop beeps and restarts after seeing a blackscreen. Surely it cant be a hardware issue if their Is more than one person experiencing this?