• Happy holidays, folks! Thanks to each and every one of you for being part of the Tom's Hardware community!

Alienware M17R4 Nvidia TDR error or dying/dead GPU or failing fan.

Frank Muszka

Commendable
Mar 30, 2016
2
0
1,510
Hello
My name is Frank, and this is my 1st post here, so if I'm doing anything wrong please let me know.
I have an Alienware laptop M17R4, with the following specs:
i7-3740QM CPU @2.7GHz
4x4GB DDR3@1600 RAM
Nvidia Geforce GTX 680M GPU
2nd GPU is an Intel HD Graphics 4000

I would like 2nd opinions/advices on the following.

My laptop would turn 3 y.o. in May, bought it brand new.
about 3 weeks ago, I started having a "Display Driver has Stopped and Recovered", since I had this error some years ago, and it was a mere driver update I didn't pay too much attention to it, but as it started to pop up more and more, I finally decided it's time to update my video drivers (no too much of a fan of PC self updating itself, rather doing it myself from time to time)
Ran like 2-3 days with the new driver but the errors didn't really stop, could still play so I thought it's something fishy with the software, one evening I started having white flashes while playing a game (diablo 3) and lag-like graphic spikes. Not too often, only a few in a span of 2-3 hours. Then in the next hour these glitches/flashes aggravated until the point the game crashed. I still thought it's only software side at this point.
Ever since I was trying to figure out what was the source of the error. Following different YouTube guides/solution, or stuff I read on different forums, asked friends.
My early thoughts were faulty drivers. So I tried all the drivers on the Nvidia site, Using DDU on each. Tried reinstalling windows, getting fatory drivers, registry edits, for TDR things. No avail.
Somewhere in the midst I started leaning towards hardware failure. While at the moment I almost certain of it, I would still like some thoughts/opinions on it.

Will try to go into detail on some of my tries.
Some stranger results include:
To eliminate the possibility of a harmful software/virus/malware, I removed my 2ndary HDD and left only a 128GB SSD, went full factory thing, with the Win7 that came with it, including some of the basic drivers which were a must for running, disabled updates from the getgo, to keep it as clean as possible. After I used an older 355.97 driver for Nvidia I managed to fun Kombustor stress test for 1-2hrs on 1280x760 and 1920x108 (stopped around 2hrs on 1st try and stopped around 1, 1.5 hr on my 2nd), without any error. Thought I gonna go further, in hopes of some better results. I installed battlenet application which is needed for most of blizzard's game, and as soon as the install finished display driver started crashing, without even tring to install any game at all.

Upgraded to win 10 afterwards on the suggestion of a friend, and by mistake I installed new newest driver from Nvidia's website. Since I had 2 GPUs I had to turn the GeForce one off, since it was constantly crashing with the newest driver without any activity whatsoever. Got a DDU, reinstalled the older 355.97, and ran Kombustor again. Was working again w/o errors. So I wanted to try some other games. Did install HoMM V (1st thing I found). Game launches, no problems in the menu, but game freezes randomly when ingame (sometimes after 10 sec, sometimes after 2-3 min). No error msg at all. to eliminate the possibility of incompatibility of an older game with win 10, I installed DotA2. Fine in menu, but freezes after like 10 sec ingame. With the same "Display Driver stopped responding and now recovered error".

Upon more suggestions I ran FurMark, which overheated my GPU 89C, and crashes it after 1 min-ish. Kombustor never yielded any higher heat lvl than 84C, that was on Win7 too. FurMark pushes it to 89 (that's the Nvidia limitation according the Kombustor's GPU shark) after 30 sec, and after a 30 sec struggle, it crashes the hole thing, with the ominous error, while crashing FurMark in the process too.

I am still puzzled but inclined to believe that the GPU is dead or dying. If anyone could confirm/deny this or knows some other testing methods that would save me some money time, since I'm from EU and it's harder to get spareparts to Alienware.
Or could it be just some bad luck with drivers, or incompatibility with current directX or whattanot?
Though I can hear both the CPUs and the GPUs fan running normally, can it be that it's failing? Is there some ways to test those?

With best Regards
Frank