Computer Reboot After Extended Play Sessions

xevo7x

Reputable
Apr 22, 2015
10
0
4,510
After playing graphically demanding games for extended play sessions, I will experience very minor artifacting then immediately my computer will reboot with no blue screen. I attempted to test why and I noticed it only happens when my CPU reaches around 75-80C and my GPU reaching roughly 65-70C. These temperatures only get this high playing very demanding games at their max settings.

My PC specs:
MSI z97 sli krait edition Motherboard
Nvidia GTX 970
Intel i5 4690k w/ stock cooler
2x 4GB Corsair RAM
850w modular power supply

When I researched online I saw my temperatures are not dangerously hot, so I'm extremely confused as all parts in my PC are fairly new and I have not been able to find another person with a similar problem. Any help would be greatly appreciated.

 

AnPCtester

Reputable
Apr 17, 2015
242
0
4,760
What is your PSU brand? How old is it? Can you check if everything is plugged in correctly?

It seems to me either you have a PSU related problem or there is a thermal shutoff limit set in your mobo bios which is turning your pc off.

just my two cents
 

xevo7x

Reputable
Apr 22, 2015
10
0
4,510
My PSU is a Thermaltake smart M850W, brand new I just got it a few months ago, all parts are plugged in correctly. I will fiddle around with my BIOS to see if such a setting exists but I have looked into my bios enough to think a setting like that would not be there.
 

AnPCtester

Reputable
Apr 17, 2015
242
0
4,760


The Thermaltake Smart M850W is a tier 3 PSU that is not to be used for gaming and/or overclocking rigs according to this extremely dependable and extensively researched articlehttp://www.tomshardware.com/forum/id-2547993/psu-tier-list.htmlhttp://www.tomshardware.com/forum/id-2547993/psu-tier-list.html by reputed experts on PSU.

I am almost sure that the PSU is not supplying consistent power(very important in gaming) to the gpu and thus putting strain on the gpu causing it to artifice and then shut down. You will have to replace the PSU.

Try to get a Tier 1 PSU recommended in the article linked. Seasonic has the best reliability and stability. XFX and EVGA are also good brands.
 
Actally the Thermaltake Smart Series units are Tier 4 units which isn't quite "replace immediately", but it's pretty bad. We see a lot of issues on builds using these units. I'd highly recommend replacing the unit with a Tier 1 or 2 unit as mentioned.

Also, just FYI, the Tier list is actually the work of one member, Dottorent, and is based off professional reviews conducted by power supply testing sites like Jonnyguru and a few others. It's not an extensively researched article by PSU experts, but it is the culmination of extensive data gathering by that member and is about the best general guide to PSU tiering that I'm aware of.

Any unit you're in doubt about should be googled for results leading to a dependable professional review, don't ever listen to reviews at Newegg, Amazon or somebody's "it works for me" word of mouth.
 

AnPCtester

Reputable
Apr 17, 2015
242
0
4,760


Thank you very much. I stand corrected and am very impressed that a single person had the dedication to do that good a job.
 

xevo7x

Reputable
Apr 22, 2015
10
0
4,510
Well this is more than irritating, Geek squad said this PSU was going to be great for my PC. I would like to be certain that my PSU is to blame before I spend more money on PC parts. I will also provide the information that my GPU can be at temperatures up to 80C when playing games like Metro Last Light and have no problems, my PC only will reboot when playing CPU intensive games at their max settings like Arma 3 or GTA V. I can also play games with no problems at all for days as long as I don't use max settings.
 
I'd do a clean installation of the drivers, as outlined at the following link in the clean drivers section, and then install the new drivers released by NVidia a couple of days ago. Even if you think the drivers are ok, I'd still do it just to eliminate the possibility of it being driver related.

http://www.tomshardware.com/faq/id-2492424/laymans-simple-guide-solving-graphics-card-issues.html


Then install the latest NVidia drivers for your card and OS type. Keep in mind that the new NVidia drivers are different for 32 and 64 bit OS whereas they were using the same driver for a while there.

http://www.nvidia.com/Download/index.aspx?lang=en-us



 
Artifacting is ALMOST always GPU or motherboard related. It's pretty unusual for the CPU or memory to have an effect resulting in artifacting that doesn't simply cause a freeze, bsod or reboot. It IS possible though, don't get me wrong. Just not AS likely.


Considering you probably have some thermal throttling going on, which may or may not be unrelated to the artifacting is a concern. I see that you have a stock cpu cooler, but I don't think it's the primary issue. What is your case model and have many of the fan locations are populated and in what configuration. Even with the stock cooler, unless you have the CPU overclocked, there is no way it should be anywhere near that hot, ever.
 
There's your problem. you have NO intake fans. Having exhaust fans is all well and good, but without any intake fans to bring in fresh air, you're relying on the pressure differential of a negative pressure exhaust configuration to bring cool air in, which is extremely poor in both theory AND application.

I'd get and install a front 120mm fan as intake and turn the side fan around as an intake as well, since that's what it's supposed to be anyhow. Side fans are intended to bring cool air in directly onto the graphics card and CPU cooler, while front intakes are intended to supply the cooler ambient air necessary for a beneficial exchange of internal heated air for cooler external air. With a rear and top exhaust and a side and front intake, your temps should drop considerably.
 
Your case has:

FRONT, 1 x 120mm
SIDE, 2 x 120mm
TOP, 2 x 140/120mm
REAR, 1 x 120mm, 23db/42CFM (included)

So I'd HIGHLY recommend using the single front location for intake, getting two 140mm fans for the two top exhaust locations, either one OR two side 120mm fans (One is probably fine but two certainly won't hurt) and of course the rear 120mm fan should be configured as exhaust. Having a good exchange of air in the case is actually even better than simply taking the side cover off as removing the side cover DOES allow more cool air to enter the case for use by the GPU and CPU fans, but it does NOT facilitate an exchange of the heated air inside the case nearly as well as a directionally oriented airflow path as used by most cases does.
 
Unless the fans that are currently intake are located on the side, bottom or front of the case, then no, you do NOT want them as intake. Rear and top fans, like I said, should be exhaust. Front, side and bottom fans should be intake. The only way that should ever change is if you have a top mounted power supply or have a liquid cooling radiator mounted in the top of the case.

The two side fans can be intake, yes, but you should STILL have at least one front intake fan in addition to those, or two front intake fans if one or zero side fans are used as intake.
 

xevo7x

Reputable
Apr 22, 2015
10
0
4,510
Of course I will keep the top and the back for exhaust. When looking at my case, the only area that I could see where a front fan would go is near the bottom, and would only blow onto the PSU and a few wires, would it still keep my pc cooler even though it's not directly blowing on my GPU & CPU?