M17xR3 Freezing while gaming with no BSOD

Hello all!

For the past few months, I have started having trouble with my M17x R3.

Around 2 or 3 months ago I began having an issue where while playing certain games (BF3, DOTA 2, Company of Heroes, Homeworld 2 are some examples) my screen would turn a solid color (no certain color. Have seen blue, white and just a black screen) and the audio will freeze and buzz. The PC will not respond and requires a hard reset.

Now this does not happen on every game and is random in occurence. The games it does happen on are not all GPU intensive. I even have some games that have never had a crash occur. (Total War: Shogun 2) Sometimes the freeze will happen as soon as I start a certain game, and sometimes will never happen for two days then twice after a reboot. VERY FRUSTRATING

I have had problems with it in the past where it would BSOD and Dell ended up replacing both the motherboard and the GPU (Nvidia 580m). After the replacement, it worked without issue for around 6 months and then this began happening. I was hoping you guys could help me out and also answer a few questions I have.

A few things I have tried:

1. Preformed a clean install of my graphic drivers. Did not help. I even went back to the first drivers I installed after Reinstalling Windows during the troubleshooting for the first problem.

2. SFC/ scannow: no errors, did not help

3. System Restoring to a point that I beleive was before the crashed began. Did not help.

4. Updating my Direct X. Did not help

5. Overheating: I thought it might be overheating but after buying a laptop cooler and making sure temps do not go over 71C and it still crashes I don't beleive it is.

Another strange thing with this problem is there is no BSOD so there is no dump file or crash report. So I am having a hard time even beginning to figure out what it could be. I looked at the event viewer and there were no warning/critical events other then the shutdown when I have to hard reset it.

I have never Overclocked or messed with any part of the system. Just wanted to put it out there for your information.

I have not contacted Dell Tech. Support yet and will propablly do so over Thanksgiving break if the problem persists and a solution cannot be found. Now I am willing to send it in for repairs if need of have them fix it at my home but I have one worry. Since the problem seems VERY similar to the first problem I had with resulted in a replacement GPU and Motherboard I am kind of expecting something along those lines again but I could be proven wrong. I have heard though that they discontinued the 580 series of Nvidia cards. Does Dell still have any left and if not, what will they do? I have 2 years of warranty coverage left and I don't think they could leave me without a card? Would I get bumped up to a higher model?

Thanks guys! Any help or ideas on what to do next would be great

-Matt

P.S: Since I just joined the forums and did not see that there was an Alienware area of the forum, can someone tell me how to delete a thread I put in the general Dell area? Thanks

 

Welcome to the forums. This is the proper sub-forum to post in. That way you get help specific to your machine. Now onto your problem:

So, have you tried running the PSA and see what it says? PSA stands for pre-boot systems assessment. To access it, hit F12 from POST (that is initial startup with the Alien head logo) to bring up the list of startup options. When the list is displayed, select ePSA or Diagnostics...it's always the last option on the bottom of the list. Next, the notebook will run thru some basic tests to check the CPU and motherboard. Let it run thru all these tests and follow any instructions it gives you. If it doesn't find any errors and asks if you want to run the extended memory tests, select yes. These next set of tests checks the RAM thoroughly for errors. It will take a bit of time depending on how much RAM you have installed. If there are no errors after than, then you can restart into windows. PSA is meant to check for hardware errors, it does not check the hard drive partitions or data for any problems. One other place you can check for errors is in the reliability history. In start menu, type reliability into the search box. Then select view reliability history. This will categorize and warnings or errors by date, with the newest first. See if there are any errors from the last time your system froze.

If they cannot replace your 580m card with the same model, then yes, they will most likely give you a new/better GPU.

PS: Never use system restore. It's like Windows Cancer. It keeps all the new problems while restoring any old ones you previously had.

 

Hi Radji,

Thanks for responding to my topic!

I ran the PSA and let it do the basic tests and the long memory test and it found no problems with the system.

I check the reliability history and the only thing it shows it the warning that "windows was not properly shut down" due to me having to hard reset it.

The only thing I did notice that on a few instances when this was first beginning, there was an error for LiveKernalEvent with a BCCode of 117. During these events the screen would turn a purple color and windows would say "your display driver had stopped responding and has recovered". These did not lead to the black screen or the hard reboot.

Thanks for the tip about system restore! I'll avoid using that!

Any ideas of where to go from here? Should I get on the phone with Dell or try some other solutions?

Please don't get the idea that I'm trying to race to Dell blaming them and asking for new hardware. The problem is that I am away at college and would have a hard time getting anything done on campus with its location and the terrible phone service I have. Thanksgiving week is the only shot I have until Christmas break to get something done

Thanks,
-Matt

 

Post your system specs so we can go from there. Its better to know all that information ahead of time. For reference, do you have Nvidia Optimus installed in your R3 or are your using the standard graphics switching method (Fn+F7)? Or do you have the 3D screen package?

 

System Specs:

Intel I7-2820QM
1920x1080 screen (32bit)(60hz) LCD
Nvidia GTX 580m
16GB RAM at 1333Mhz
OS: Windows 7 Home Premium 64-bit

I do not have Optimus installed. When my GPU and Motherboard were replaced last time, I was told by the tech. that the Intel HD 3000 Graphics would not be recognized by the PC. I have not used or seen the Interl HD 3000 on my PC's device manager.

I do not have the 3D Screen Package.

 

Hmm. Interesting. Sounds like your BIOS is set to PEG instead of SG (PCI express graphics instead of Switchable Graphics). I didn't think that could be done without an unlocked BIOS and I know Alienware wouldn't have a tech install a motherboard with unlocked BIOS flashed on it. Restart and go into your BIOS. See if you have a tab labeled Advanced and if there is a menu option called Video Configuration. If it's there, select it and see what it says under Primary Display. It should say something like PEG or SG. I just want to determine what your BIOS is set to with respect to the GPUs. Another thing to go for is to download GPU-z. Its tells you all sorts of details about your GPU. Open GPU-z and see what it tells you about your Nvidia GPU's BIOS version. The current version should be 70.24.2B.00.06. If it's not, download and flash the newer VBIOS here: Drivers and Downloads | Dell [United States]

 

I downloaded GPU-Z and the BIOS version is the correct version (70.24.2B.00.06)

I was able to get into the BIOS but was not able to find the Advanced Tab. While I was not able to find that, as I was getting ready to log out I happened to get to a screen that gave the option of restoring to the default settings. I restored to default settings and now my Intel GPU is showing up on my system.

I'm not sure if that is the info you wanted so if you need anything else let me know and I'll try to find it!

 

Not really what I was hoping for but I will take it. The issue at heart is why your system is suddenly crashing. Restoring the BIOS to default is a good place to start. But remember, setting it to default means you will have to switch GPUs when you want to play any graphics intensive games. One thing to consider is to also try using the Intel GPU for one game (preferably one that won't kill the Intel GPU) and see if you still get the system freeze.

 

I was thinking of that. Well I know one game that I play that always crashes while running the 580m. I'm going to try to play it tomorrow with both cards for around an hour or two and see if it only crashes on one or both. Hopefully doing that will give us a better idea of what's going on. I'll try some other games too to see what happens. I'll report back with the results!

 

I had exactly this happen on my R3/580m. I monitored the temps in game and saw it happened at around 5 seconds after GPU temp hit 72C. Weird, - lockup and buzzing. Dell replaced the whole machine with this r4 so no idea what it was - they just gave up! Good luck.

 

Mickey,

Thanks for the post! Glad to hear that Dell took care of your problem successfully. Gives me some hope for mine! How long did you have your R3 when they replaced your system? I've had mine for over a year so I'm not holding out hope for a whole new system. Gotta admit though. For the thing to have a major problem like this twice in only a little over a year of me owning it is slightly annoying. But, I've read about people having a much worse time.

Hopefully tonight I'll have a better idea of what is causing the crash. If it runs on the integrated fine but crashes on the 580m then it would be the exact same symptoms of my GPU/MOBO needing to be replaced like it was last time. Just like yours though, they never figured out exactly what caused it.

 

You're very welcome - I feel your pain!

I'd had mine about 6 months but I've read about an R2 owner with a 3 or 4 year warranty getting a complete replacement (an R4 so arguably a downgrade since he had a 280 sli and got a 675m but comparable I would guess) so length of time since purchase does not seem to matter.

I had a HD failure first, then a bunch of problems with the 580 overheating. Replaced heatsink, then the card started to artefact, probably, they said, it had been 'cooked' too many times so they replaced that AND the heatsink again. Finally the lockup and buzzing started after that. When they suggested the next thing was a mobo I freaked (well more of a 'I'm so disappointed with Dell, paid all this money and not got to use it properly for 6 months!) and asked if there was any chance of a replacement? He said he'd check and came back with 'you've had 3 engineer visits so you qualify if the senior agrees' - he did! They upgraded the CPU as well as a bigger second HD and a killer wireless I never had. To top that they gave me a new warranty and accidental cover from the date of delivery of the replacement - top marks and a very happy customer

 

Ouch seems like you had some problems too. So far I've had 3 visits. The first time was due to the pc being shipped with a defective keyboard. The tech tried to fix it the first visit but Dell sent a keyboard in the mail and it was broken in shipping so he had to come for a 2nd visit. Third visit was for the GPU/MOBO replacement. After I call this weekend and we figure out the next step, I'll give them one more shot to fix this problem but I think if after a 2nd new GPU I have this problem again I might ask them what options I have available. Granted they are very helpful when I call, I don't want to have to do this every 6 months to a year til my warranty runs out

 

Good luck with it, main thing I learned is to try to get empathy from the tech, not attack. I'd be inclined to try the 'this is getting me down, any chance of a replacement?' on your next call. Even if they say they would like to keep on trying you have made the first move (and the tech gets to close a tricky case so a plus for him as well!)

 

Thanks! Yeah I always try to be as respectable as possible to the tech. They are just doing their job.

Well I've been trying out a couple games while using the integrated graphics and there has not been one crash while playing. This makes me beleive that the culprit is the 580m. I'm on close to 4 hours of gameplay with not one problem while the 580m has been crashing after 15-30 minutes and more frequently as days go by.

Radji if you read this any ideas? Anyone else have any thoughts on my belief that it is the GPU?

 

Other than another dying 580m card? Nope. But this time, see if support is willing to do a parts only service. No tech. They may just be doing their job, but half of them can't even do it proficiently. Parts only service means they will only send out the parts to you and you send back the defective part (i.e. the 580m card). This of course is only an option if you feel comfortable enough to open up your R3 and replace the 580m card yourself. If your not, then you will have to go with the tech again (or depot service...whichever they offer) and hope it gets done right. Replacing the GPU yourself isn't terribly complex, but if your not familiar enough with taking the machine apart, then that avenue is probably not for y'all. I only suggest it because by doing the work yourself, you can take precautions to make sure the GPU is installed right (and to limit the instance of ESD).

 

Thanks for the idea Radji but I really don't feel comfortable taking my system apart. Luckily, where I live the tech service is really good and every time they came out they were very professional and explained everything they did which put me at ease. Hopefully they will do an inhome instead of depot though. Heard negative things about that. Thanks for the imput!!

 

Hey all!

Sorry for the late reply but finals and photo-finish papers were taking over my life for the past few weeks but I wanted to give you an update! So... I was able to call Dell over Thanksgiving and they went though numerous tests and things to try to fix the problem. Along with the various ideas everyone gave me on this site, we:

Updated the System BIOS to A12
Made sure that the GPU drivers were the most recent ones and did a clean install of them.
Fiddled with some settings in the Nvidia Control pannel

None of those successfully fixed the problem. I will say though that when we reinstalled the most recent drivers, there were no crashes for 5 days but they startd back up and even more frequently.

Dell's next idea was to do a full reinstall of Windows which I am fine with doing. My question is after a clean install of the whole operating system and wiping the harddrive, can there be anything other than a hardware problem causing this?

Thanks,
Matt

 

Only the motherboard and GPU. BSODs are usually (like 99% of the time) software or driver issues. The only time a BSOD is hardware related is if the hardware component momentarily fails causing its driver to momentarily fail. But that is the only instance I've seen with regards to both RAM, GPU, and hard drive. YOU on the other hand are not getting any BSOD so that points more towards a total hardware component failure since the system apparently doesn't have time to throw out a BSOD. If a clean install of Windows doesn't work, then replacing the GPU or motherboard may be the next step. The trick is you can test the CPU, motherboard/chipset, and RAM functionality with the PSA diagnostics. It does a reasonable good job of finding any hardware malfunctions with those components. Not so lucky for the GPU. ePSA doesn't check that. If you try running 3Dmark benchtest and it causes the same shutdown, that is a big red flag that your GPU is more the culprit than your motherboard. That's how I found out my GPU was terminal.

 

Radji,

Thanks for imput! One thing I did forget to mention which I feel dumb for is around two weeks ago the PC started to actually throw out a Blue Screen instead of just freezing and becoming unresponsive. The Blue screen comes out as a Video_TDR_Error with a 0x116 bugcheck code. I did let Dell know this and that's when they recommended the clean install of windows. It points to the nvkflt.sys which if I'm correct is a part of the Nvidia drivers which makes me hold out hope that it could be something software related that the clean install will fix! Sorry for making you write that out and then throwing this info at you!

One other piece of info that I found is that while running the gaming test on my 580m from the Alien Autopsy program that the Dell tech support agent installed to test the system, the crash sometimes occurs on a test called the Moduo X test. I'm not sure if you know anything about what that test does but hey, any info I can get I'll give!

Sorry for the late info Radji!

-Matt

 

I am having a similar problem. I have an M17x R3, 580M, Samsung 830 SSD, 16GB Ram, Windows 7.

My computer will white screen crash without any errors and it requires a hard reboot. It happens at random. It can crash just from surfing the web and reading basic websites. The vc temp never goes above 65. It passes all the hardware tests and 3D benchmarks. At random the screen goes full bright white and it is non-reponsive. The video card fan rolls to high speed and the machine gets very hot until I shut it off.

I have had a series of errors from kernel 41 to BBcode 117. The white screen crash does not create a crash dump report. I have cleaned and installed the Nvidia drivers & Dell's AW drivers including the original ones (10 months old) to the latest ones. Dell recently replaced the motherboard and it didn't fix the problem. Today they replaced the 580M. It will take a few days to see if it crashes because it is random.

I am curious how many people get their hybrid graphics to work. I wonder if my problems with the hybrid indicate a bigger problem. I have several laptops with hybrid graphics and they all switch without any problems, except the M17 does not. I have the latest BIOS (A12) and I have to select PEG in the BIOS to get the 580M to work. If I use SG in the BIOS the Intel graphics will always be on. They both show up in the Device Manager. I have changed all the power settings, even the PCIE settings, Nvidia settings, and Intel settings. It's not a big deal if the integrated graphics don't work. 95% of the time I am plugged into power, but for a laptop this expensive it is supposed to work.

If it happens again, I know Dell will tell me to reinstall Windows and that's fine, but I doubt it will fix this. I like the machine and Dell's support has been good. I am tired of spending hours trying to fix this on my own and I can't sell it to someone else in good faith that it will work. I will post back in a few days if the 580M replacement worked.

 

(Warning: Follow any ideas I have at your own risk. I am not certified in anyway! Just want to throw that out there!)

Osiris,

Thanks for posting! I'm no M17x expert but I think I might be able to give you some idea for the problems based on my own experience:

Problem #1: The White Screen-
This sounds exactly what was happening early on in the issue I am having currently. Does it happen at all while gaming? Mine would rarely happen while on the web but more so when playing games. After a while of throwing a random colored or a black screen, it finally started showing the BSOD with 0x116 codes. For that problem i still have no answer but hopefully will soon!

Problem #2: The hybrid graphics-
Do you have the Intel HD 3000 in your machine? For myself, all I had to do to get the integrated working was go into the BIOS and reset it to default settings. I'm not sure if you fiddle with your BIOS and wouldn't want it back at default but it's worth a try! Also did you try changing the NVidia 3D settings to prefer using the high-preformance processor in the control pannel?

Thanks for the post! Hope you get your problem fixed as well!
-Matt

 

Matt:

Thanks for the help.

Yes, mine will crash from anything. I do tab out of a lot of games to check on the web, but that doesn't consistently create a problem. Sometimes I can play a game, then exit the game and surf the web, then it crashes. It is exactly like this video, but mine is totally random. If it crashes again, I am going to try his approach about hitting the FN HDMI switch. That's an interesting idea.

Alienware M18x Laptop White Screen *FIX* - YouTube

I do have the Intel 3000 graphics. They are turned off in the BIOS and the Intel video drivers do not show up and neither does the Intel card in Windows Task Bar & Device Manager. I read last night that the Alienware's rarely work with the Intel and Nvidia graphics.

Several months ago I was getting the blue screen crashes. Those stopped when Dell gave me the A08 BIOS unlocked and I turned off the Intel graphics. Now I just get the white screen crashes.

Yesterday Dell replaced the 580M and I have not used it enough to know if that is the right fix. I did more research and it appears that this problem happens with Nvidia and ATI cards. It reminds of all the video problems with DX10 and Vista when it was new. I would put in another drive and run Ubuntu on it, but I don't have any good games for Ubuntu. I am hoping that situation improves soon with Steam/Valve's support.

One odd thing that I noticed is that sometimes when it is plugged in, the laptop switches to the battery for 5 seconds (randomly) and then it comes back to the power supply. Most of the time it crashes when it switches the power supply, but not always. That's one of the reasons Dell originally replaced the motherboard. I also have two power supplies from Alienware and it crashes with both power supplies.

I will keep you posted if I have any success with the new card and if I have any crashes.

Thank you for the suggestions.

 

Well time for an update!

After talking with Dell on their facebook page, they asked me if I would be willing to do a full system reinstall. I was fine with it so I went ahead and wiped the HD and reinstalled only windows, the essential drivers and one game to test. Annnd the game gave me a bluescreen with the same error on the first startup.

I got on the phone with Alienware yesterday and after telling them that the clean install did not fix the problem they promptly sent a new 580m, a heatsink kit and even a new fan. Parts have arrived and a tech is coming out to service my machine tomorrow!

Fingers crossed! Hopefully this will get this cleared up!

I must say though, I am impressed at the speed Dell gets these parts to your home. Granted they may have to troubleshoot alot before but both times they have gotten the tech to my house in 2 days which is pretty good in my opinion! props for that Dell/Alienware!

-Matt

 

8 days ago they replaced my 580M and I have used it a fair amount. I traveled a few days so I didn't use it every day. So far it has not crashed with the new 580M. This is a long time for the laptop to go without crashing. I think the problem might be fixed.

I use my machine almost every day. If I crash any time soon with the same white screen crash, I will respond back. This has been a very painful experience with Alienware. I will say that after you get through the initial run around on the phone, Alienware has continued to work on it and they didn't leave me hanging on this.

Matt, I will cross my fingers for you. There could have been a bad run of 580Ms in these models. I hope the new card works for you.

 

Hmmm. My two year old M17x r3 with 580m started throttling last week or so. I had been running something called NVI_0_2_1240_1500.exe for all that time. I can't recall how this was related to throttling, but it supposedly helped (Anyone know?)

Well, the machine throttled on a game I'm playing, then just went to blackscreen a few days ago with a sound playing (the last sound of the game). I rebooted and all was fine.

However today the crash happened a few times. I ran Furmark, and all was fine until the machine throttled after about 8 mins. It then throttled again a minute or two later.When I looked back shortly after the machine has blackscreened. I rebooted and re-ran Furmark. This time it ran about 30 seconds before killing the machine. I didn't see any sort of throttling....

I haven't stressed the card since, and the machine seems to run fine. I should point out I've seen absolutely no artifacts when running the card generally.

My questions:-

What should my course of action be? I'm going to buy a laptop cooler tomorrow to see if that helps, but should I be looking at buying some sort of thermal paste? (Never done it before) Should I use HWInfo to run the fans at full speed? Or even replacing the 580m with another or a 680m? The machine has passed the F12 diagnostic tests, the fans seem OK, Alien Autopsy tests all pass etc etc.

Cheers!

 

Hi there. The 580m is a poor card from my experiences and research. When I first got my R3 throttling was a BIG issue for a lot of people. Just take a look at this enormous thread: http://forum.notebookreview.com/ali...x-580m-78c-throttling-see-post-191-later.html

Like a lot of problems it's a combination of events making it difficult to positively identify the cause of a 580m issue. As osiris7374 said, The lockups and driver stopped messages is exactly the same as I and many others experienced when MS released Vista. AMD had the same issues. Root cause was the way MS had changed the HAL and forced code to go through it rather than be able to directly access the hardware. This would not be a problem for any driver author as long as they stuck to the MS published API's and calls. Video OEM's write their code to get through this layer as quickly as possible and often use undocumented calls - so they could not blame MS.

Well, on the release of Win 8 the same problems started. I was sure it was the same thing but basically got trashed on the forums for suggesting this was the cause so I just watched. Seems time has proved me out correct and some people even found the windows 7 update (for win 8 compatibility I assume) that triggered it. NV cannot publicly blame MS since they did it to themselves and trashing MS is not a good move for a company that relies on developer status and access.

Now the stars start to align (badly!). NV introduced their 'boost' feature with the 3xx series. I saw that this was affecting my card (no boost support on this model) by causing a brief voltage hike to 0.92v. The reference voltage being 0.87v. I had already seen people that tried an overvolt vbios to 0.92v have their cards fry! It never seemed to cause me any problems until 320.18 was released. See this thread for details: http://forum.notebookreview.com/alienware/721878-nvidia-320-18-driver-warning.html . Cards were dropping like flies and factory overclocked desktop cards were the most affected, clearly having a meltdown!

Then add in the fact that NV couldn't get a driver out that worked good with their latest cards, this is still the case with their latest attempt, and you have a resource problem they have struggled with. Along with their investment and focus on a handheld device and streaming a PC game to the handheld IMO they stretched themselves too far not anticipating all these driver problems.

Upshot, you might have a simple problem of the heatsink clogged, have you cleaned it out? A driver conflict with your setup, use the dell published one, OR, the card is failing and only replacing it will show for sure - though and it's an expensive fix if not needed. The 680m has been a very successful upgrade for the R3 but research it, it's not straightforward and you have to disable the on-board compromising battery life (and the environment ).

Good luck and ask away if those threads raise any questions

Edit: Forgot to add, MS have published a new set of calls and default actions specifically for Video OEM's: TDR changes in Windows 8 (Windows Drivers)

Looks to me like all NV have managed to implement on their latest attempt is the 'driver stopped responding' message not appearing. The pauses and stuttering remain

 

Thanks for the reply Micky.

I've not changed drivers in absolutely yonks, so am not sure if that would be an issue.

I'll try cleaning the GPU etc if I can find a proper tutorial. I have a PSU arriving tomorrow (the blue light on mine went out. It doesn't come on, even if not attached to the M17). I'm hoping that this is causing the issue, and that no lasting damage has been done!

Cheers!

 

You're very welcome .

If you haven't changed the driver in ages then that is an unlikely cause. Might be worth just updating to the latest one dell publish for your service tag : Product Selection | Dell UK

You might find that just cleaning out the heat exchanger is all that is needed. No need to open up the card and repaste/pad it. Just remove the fan and the two screws holding down the card and heatsink. The first time I removed my assembly it was like the black hole of Mumbai in there .

Edit: more afterthoughts! It has been known that a PSU problem can cause the card to drop to a lower power setting - might be the cause also

 

Forgive my ignorance, but what is the heat exchanger?

I was looking to remove the fan as discussed on page 32 here:-
ftp://ftp.dell.com/Manuals/all-prod...ps/alienware-m17x-r3_service manual_en-us.pdf

Did you mean take the fan and the card out, and the heat exchange is there?
https://www.youtube.com/watch?v=6vEk_ZFAzLI
3:20

The actual GPU inlet doesn't seem to take that much air in (though it pushes a fair amount out) so perhaps it just needs a good clean.

 

OOps sorry engineering technobable! The bit at the end with all the cooling fins on it that is in front of the grill outlet.

That doc is perfect. Next Just follow page 35 to remove complete unit with the the gubbins attached

Let's hope a good clean is all it needs. I got some cans of compressed air for all the cleaning. Just be aware that if you buy this, the cheaper ones often cause freezing residue if you use then too much. Not a problem as it quickly evaporates but not nice to use .


Good luck.

 

Thanks again. I'll try tonight.

Unfortunately the PSU won't arrive until later this week. I'm not that hopeful given the machine turned off about 2 mins after turning on last night , but we'll see.

 

Unless you heard the fans crank up before it turned off it is unlikely a clean will help, but sometimes just reseating the card fixes things

 

Cheers. Will see. It reached about 77-78 the other day before shutting off. I think it did the same at 71 the next time. I'm still desperately hoping that some combination of clean, paste, new PSU will fix it. My one bit of hope is that the machine seems to work fine on battery. I'll report back when done.

 

A quick update:

I received my new PSU, cleaned the CPU and GPU (didn't repaste them), popped the M17x r3 onto a laptop cooler.

Ran furmark and OCScanner for about 15 mins and 10 mins respectively. All fine.

Ran the game I'm playing, it crashed after about 20 mins. As soon as the machine rebooted I ran Furmark, which crashed the machine after about 1 min. (Game not running.)

However, the logs for GPUZ and RealTemp indicate 63 for the GPU and 52 for CPU (All RealTemp + Prime 95, Memtest86+ and disagnostic tests pass).

When I ran the game at a lower res and slightly lower details I got a full hour out of it, without crash. I turned it off as I had to go out....

So, is there any way to confirm it might be the GPU, other than try and find someone else with an M17x? £300-400 is a lot to pay on guess work (The fact the temps are so low on the crash is worrying, given the card was running at 77 a few days ago - with no laptop cooling - and throttling a bit before the machine crashed. I haven't seen any throttling since).

So, I'm running out of ideas on how to isolate the problem. It might be the motherboard I guess, but I've no idea how to test this...

 

Sorry to hear that man . Unless someone else comes up with a good idea (and I'm out now!) it really looks like a card failure now .

You've certainly removed overheating, which it appears it was doing. The first 580m I had Dell replaced the heatsink as the card was running hot. once they fixed that it still locked up occasionally. The consensus of opinion was that all the hot running had damaged it . They even put another new heatsink with a new card, i'n in case the heat had warped it'!

As a final check on the possibility of temp being an issue (and trigger) try using MSIAfterburner to give you an on-screen-display of realtime temps and clocks when gaming. Don't have any other monitoring SW loaded and active. If you find that a 10% downclock from the base setttings gives good results, but of course poorer gaming, then the card even more looks like it's left the game.

These NV cards appear to have a high failure rate and warranty is a safe bet you'll spend less - I've had three GPU's and a full replacement .

I think you're going to have to roll the dice on a new one...

 

Cheers sir. I'll give MSIAfterBurner a go.

I'll see how long I can ride it out, then decide whether to buy a new card, or a new/used system.

 

Mickey,

I see you have the 675m. Any ideas if this card would allow 3D on the M17x r3, or is it the cast that only the 580m would work? Cheers!

 

Hey Jimbob,

It will not work in 3D I'm afraid. There is some sort of 3D certification that goes on in the bios, and only cards sold with a model work. Kinda annoying since it would be a simple bios fix for dell but they never revisit an older model unless a problem is identified .

And you'd be into modified drivers - for that path (without 3D) the 680m is your best bet .

The 675m is identical to the 580m and I did read about some people flashing a 580m to a 675m (for absolutely no benefit!), but going the other way, and even then the device ID must match, it's fraught with potential issues.

The 675m is no more reliable than the 580m but I guess more easily available?

Good luck.