P770ZM shutting down suddenly while gaming

My specs:
Chassis: Clevo P771ZM
Motherboard: Intel Z97 Chipset
GPU: GeForce GTX 980M 8192MB (8GB) GDDR5
CPU: i7-4790K | Quad Core | 8 Threads | 4,0 GHz
RAM: 16GB (2x8192) SO-DIMM DDR3 RAM 1600MHz Crucial
SSD: 512GB SSD Samsung m.2 XP941
SSD2: 256GB Crucial C300
OS: Windows 10 64 bit version 1607 (build 14393.693)
NVIDIA Drivers: 378.66 (issue also happened with some slightly older drivers)

For the past few days now I am experiencing this sudden shut down while gaming, or in one case, while in Windows but with a game minimized, running in the background.

It doesn't give me a blue screen of death, the video don't freeze and the audio doesn't loop right before the shut down, there's no system sound, it just shuts down, doesn't restart automatically.

So far I've experienced this is two fairly graphically intense games, Batman Arkham Knight and Resident Evil 7. The shut downs don't seem to happen when a lot of computing power is not needed, for example if I'm playing some other game that is not as demanding.

However, just to make it clear, this issue has appeared a few days ago and it's not about my machine not being able to handle these two games or heavy games in general, since I had been playing them for quite a while and everything was okay, even the frame rate was almost always 60.

So the first thing I did was to update the drivers, that didn't do anything of course.
Then I remembered my computer has a slight undervolt done in an attempt to reduce temperatures a bit. So I restored the values to default, shut down still happened.

I have run HWiNFO64 before and during the last shut downs while playing Resident Evil 7. I was actually recording the playthrough for youtube with OBS, but crashes have happened even without OBS open. Anyways I'm attaching a .zip archive with these two logs. If you open the files with Generic Log Viewer 3.1 you can see graphs and compare them, it's super useful. Otherwise you have to read them in form of spreadsheet which is quite a mess to read.

Anyways I am not a super expert so I can't really detect the cause of the problem by looking at those logs, although I notice that both virtual memory and physical memory load and usage seem to go to 100% for a while right before the shut down. Other values are also too high but they are not consistently at the max for a while right before the shut down.

With that said, I am now at a loss for possible causes and solutions, while at the same time I feel already very worried about the health of my computer (which just turned 2 years old) and stressed that I can't even do some gaming after work now, nor follow my planned schedule for my youtube videos (might sound stupid but yeah to me it's important).

When I have problems I always come to this forum and you guys always point me in the right direction and teach me many things, so I hope together we can solve this problem as well. I'm in your hand guys.

 

Simple: Stop gaming.

What are your temps while playing the games you mentioned?

 

Not good... Looks like maybe a thermal shutdown. Even if something else is causing it, this needs to be fixed... pronto.



Edit: Oh, yeah... checked the second log file and the GPU is cooking. You need to stop using your laptop until you rectify this. Intel CPUs are almost bullet-proof, but that's going to kill your GPU. Have a look at the log graph.

 

As the log shows, they are high, but they are usually that high with any demanding game.
In this picture you can see one of the various values for the CPU temperature in a timeline, as well as 3 other values (among many) recorded until the shut down.
You can see my FPS (almost always 60, so the performance is good and the machine is not struggling), and the physical and virtual memory which goes 100% for a while before the shut down, I don't know if that's normal but all other values didn't seem to have this behaviour.

If you download the viewer and the logs, you can analyze and compare many of these values, maybe your experienced eye spots something unusual?

EDIT:
Oh thanks for looking into it Mr. Fox, I was still writing my reply to Prostar while you posted yours. Yes the temperatures are high but as far as I can tell they have always been so high every since I got this laptop, I even got into a polite feud with mySN, from whom I bought the laptop from, when I first got it (because of a performance issue that was later solved), since they were claiming everything was within what is considered expected behaviour (including the temps).
Also like I said, I only started getting this behaviour a few days ago, I have played both of these games for a while at the same settings with no problems at all, the shutting down situation has been a sudden thing that started happening last few days.

 

Ouch! That CPU must be sweating like a sinner on a Sunday. Thermal shutdown looks very plausible.

 

Yeah, you're in the process of destroying your video card. You're not seeing the most important piece of the data in the graph you shared, which is the GPU temperature. Those temps are unacceptable and your system is turning itself off to keep for bursting into flames. Stop using your laptop until you clean out the heat sink radiators and fans and apply new thermal paste. That probably won't damage your CPU, but I can guarantee you that if you leave it like that and keep gaming with temps that high your 980M is headed for the last round-up. 100°C is basically the thermal trip point for 980M. You need temps in the mid-80's or less. Your performance will also improve once you fix it because the CPU and GPU are both thermal throttling trying to stay cool.

Edit: if you are really lucky, the heat sink radiators are plugged up with dusty bunnies and lint and merely removing the fans and using a vacuum cleaner or compressed air to clean up the fans and radiators will be enough to fix it. If they are already nice and clean and that does not help, then you will need to remove the heat sink assembly and re-paste. The good new is, that is super easy on the machine you have. It won't take much time at all and it's easy to access.

 

I removed some dust from the fans just yesterday, but I guess I'll have another go and clean deeper from the heat sink radiators which I didn't really clean yet.
Right, so my temperatures are super high, actually I just checked some older logs from when everything was working fine, with a game also graphically demanding, Dying Light, and I can't see temperatures so high for both CPU and GPU.
I believe the temperatures being so high are a symptom of the problem rather than the cause of the problem. What I mean is, I think a few days ago, when everything was working fine, the temperatures weren't so high. Something must be causing them to rise like this.
When I first got the computer I was experiencing poor performance and later it was discovered I had a malware mining for bitcoins which was affecting the whole system. Once that was gotten rid of, temps went down and performance went up. In this case performance is not affected though, but temps apparently are, so I'm wondering if again my computer is secretly running something that is causing the system to overload.

I think the focus should be to discover what has changed in my system since a few days ago, when everything worked fine. I could do all the steps like applying new thermal paste, but the cause of the problem would still be there, I would just be mending the results of it. If I can find the problem instead, temps should go back to normal.

EDIT: Okay, see, here there's two graphs compared. The RED line was recorded while running XTU's (Intel Extreme Tuning Utilty) stress test after doing a light undervolt, in June 2015, while the GREEN is while playing Resident Evil 7. You can see my CPU temps are always high, and I always see them that high, but the GPU temp being so high is new.

 

@ignorant yes you could have some malware issue causing high CPU usage. Not sure it would do that with the GPU, but it might... especially if it is something related to bitcoin mining.

At any rate, your temps should not be this high even if malware is causing it. If the heat sink radiators and fans are already clean, the only solution to the thermal issue is to repaste. It's so easy to do, there should be no hesitation.

 

Yes, the GPU temp is probably the cause of the thermal shutdowns, and it's the problem most likely to result in catastrophic failure if not corrected right away.

 

Okay so we narrowed it down to the GPU, that's a start.

So what could have affected my GPU to go crazy like this with the same games it was doing fine just a few days ago? I'm doing a system-wide antivirus search and I will try to look at each individual process running in Task Manager, but other than that, what can I look into?

BTW, a correction, the graph of the stress test in the previous pic wasn't XTU, it was Prime95, but not much difference.

 

......

Just fix what you know is a problem and then worry about the unknowns. Both thermal paste and pads have an expiry and if they haven't been changed in 2 years of running near boiling I'm almost certain they would be due.

While the GPU core can throttle to reduce its heat output (before triggering a shutdown if it continues to get hotter) other components on the GPU board can't. Memory chips overheating may cause visual artifacts in games but they, and also MOSFETs, just cause a sudden cut out when they reach their thermal trip point. It's tough to tell though, hard to measure.

Malware btw will sit in the background constantly, if your idle usage is high that's when i'd suspect it.

 

I'll test idle usage, actually, good idea.

Regarding the fixing, you mean the thermal paste? Yes I can do that, sure, but guys why I do I feel like you guys are not giving the right importance to the detail that the whole system had been running fine up until like 3 days ago and now all of a sudden this is happening. I had been playing Batman Arkham Knight for like a month or more at same settings and not a single shut down or even performance problem until 3 days ago. Resident Evil 7 I played it only in one long session before but it was the same, no problems whatsoever, performance was great. So yes, I can do the "fix" of applying the thermal paste, cleaning out fans/vents, all good stuff, but that would just be mending the consequences of something that is causing the problem, not fixing the problem.

In the past I had components failing because of high temperatures, and they would shut down, but the shut downs would be gradual throughout time, happening rarely at first and then more and more often until it's constant. I hadn't experienced this at all and now I have shut downs within 10-15 minutes of gameplay. Yes it's the temperature, agreed on that, but something has affected the temperatures in the past few days since before that they were fine.

 

You told that you have seen those temps from the beginning. Those temps are too high, thermal compound degrades and you'll see even higher temps. GPU should run in that system 60-70c and cpu 70-80 easily. Monitor your temps and change your thermal compound!

 

I highly recommend Prema BIOS and vBIOS. Fn+1 will give you max fans

 

The answer is simple: thermal paste degradation, or a defective/poor fitting heat sink (also a common problem with Clevo), or both. No need to explore anything else as the cause of the high GPU temperatures. Doing so is a waste of your time. If your GPU temps are still as high after using a high quality thermal paste like IC Diamond, then your heat sink is either defective or not fitting properly (which may be a manufacturing defect if the temps have always been higher than ideal).

You may benefit from new pads as well, but the temps are the GPU core. This is like a game of Monopoly. You will "not pass go" and will "not collect $200" until you stop and re-paste. Malware is not the underlying cause. You're GPU should not be getting that hot whether you have a malware problem or not. If it does, you would have the same problem after the malware is removed under high load conditions.

 

Sorry for the late reply guys, but I have good news!
Well I mentioned that I had already cleaned the fans, yeah, but I was stupid because I hadn't cleaned the heat sinks yet, so actually today I went and opened things up and cleaned everything properly and now, wow, you can see the difference! No more shut downs and you can see the temps are much lower. So that's solved thankfully although I think that I will still look into the thermal paste as you guys mentioned. I have never done that myself but I think it's time I learn a new thing. Even if temps are okay now I'll probably need some new thermal paste in the future anyway. So I might as well look into it now.

Which one is the best in quality, or quality per price? What do you guys use? I guess I'll have to order it online. If you have bought it online from somewhere already, feel free to share links!

Red line is Batman Arkham Knight before proper cleaning, and green is after.

 

Gelid GC-Extreme is really good and lasts long

 

Awesome! I knew you could do it. Good job.

I am a huge fan of Coollaboratory Liquid Ultra. That's what I use whenever I can. It works better than anything else and never degrades. But, it is electrically conductive and not the best option if that concerns you. Also, if the heat sink fit is sloppy, it doesn't work at all. It is not very accommodating when heat sink fit is poor. (Phobya Liquid Metal, Liquid Pro and Thermal Grizzly Conductonaut work equally well, but they are not particularly safe to use in a laptop because their watery composition is more prone to movement or dripping. Liquid Ultra is thicker and stays where you put it, so play it safe and only use Liquid Ultra in a laptop. Better safe than sorry.)

The only non-metal thermal paste that I get outstanding service from is IC Diamond. It is extremely durable and works as well as Gelid GC Extreme or Thermal Grizzly Kryonaut. Those, by the way, are the best three non-metal thermal pastes. I do not find Gelid or TGK to be nearly as durable as IC Diamond. If your heat sinks are fitting sloppy, which is a common and unfortunate problem for Clevo, IC Diamond is your best bet. It is much thicker and more effective at accommodating the sloppy fit. Gelid and TGK are not effective when the fit is sloppy because they are not thick enough.

 

@ignorant Here is A Good advice for you. Do not buy / use other thermal paste brand than the aforementioned above. Waste of time and money.