Cold reboot early in graphically intensive games?

[Goregrim]

Hi there.

I've had this issue for about 1 and a half months now. Upon either immediately entering a game, or about 5 minutes into the game, my computer will do a cold reboot. It's as though a power switch is just flicked and it immediately crashes and reboots, no BSOD's or anything like that.

I took it to a computer repair place a month ago. They said the RAM was getting old but no significant issues. They said they found a virus which they determined to be the cause of the problem. However, it continued to happen after this. I was given two 4GB sticks of RAM by them to test as opposed to my current RAM.

During this time (~3 weeks) all games ran perfectly (CSGO, Witcher 3, Dark Souls 3). I ended up giving them the RAM back, intending to buy some new RAM (as that appeared to be the issue), however I held off for 2 weeks or so. During this time my games ran perfectly with my old RAM (which was meant to be the problem).

After several weeks the cold rebooting started again so I used my warranty to pick up some new RAM (2x 4GB sticks of Kingston Fury. The cold reboots continue to happen with the new RAM. I've tried the new RAM in every slot (A1A2, A1B1, A2B2 et cetera) while making sure that the RAM is completely clicked in and the results range from giving me ~5 minutes of gameplay to crashing immediately upon entering a game.

Prior to me diagnosing it as what I thought to be a RAM issue, I'd updated drivers, done heat tests, done memtests, reset CMOS and so on. The fact that the issue seems to be intermittent (but consistent during those intermittent periods if that makes sense) has me befuddled. Also, I'd swear it happens more in the ~3pm to ~3am range but I could be nuts.

I was wondering if anyone had any thoughts? At this stage I'm thinking either PSU or Motherboard, but the computer tech said there were no issues with my PSU previously (and didn't mention anything re Motherboard). The inside has all been recently cleaned and everything is plugged in correctly.
Computer build is: Intel 4690K ASUS Z97 Pro4 G-Skill 4x2 1866MHZ Ripjaws (replaced by the Kingston Fury with same specs) Seagate Barracuda 2TB HDD Gainward 970GTX 4GB Corsair Carbide 300R Antec EDGE 80+ Gold Full Modular 650W PSU

I would greatly appreciate any thoughts!

[potis]

Cold reboot and RAM..And these people work and i dont..Yay for my country and lack of jobs

A cold reboot after Windows has booted up is either from PSU or Motherboard 99% of the time, every other component would produce a BSOD and does not have the ability to "Cold reboot".

A cold reboot before that, or during Windows loading indicates some chip on the motherboard has driver issues which failed to load during check, therefor producing the cold reboot.

Heating issues usually cold reboot also but i doubt you wouldnt have checked that as you mentioned.

So, since you are getting the cold reboot during "Games" aka during loading it highly indicates power issues.

You can start testing one by one rather easily.

PSU-->GPU-->Motherboard, test the Motherboard last cause chip issues that short dont always get reproduced therefor you might not figure out its that easily.

I only include GPU because PSU problems instantly reflect on the GPU since it is the only component that requires lots of Wattage during load and it will always trigger a bad PSU.

[Blackmist]

PSU. Possibly. Some of the hardest shit I've ever tried to diagnose has been PSU related.

[Zeta333]

PSU. Possibly. Some of the hardest shit I've ever tried to diagnose has been PSU related.

Yeh, pc turning off and not Bluescreening is either overheating or psu. Since u dont see overheating its most likely psu.

[Goregrim]

Cold reboot and RAM..And these people work and i dont..Yay for my country and lack of jobs

A cold reboot after Windows has booted up is either from PSU or Motherboard 99% of the time, every other component would produce a BSOD and does not have the ability to "Cold reboot".

A cold reboot before that, or during Windows loading indicates some chip on the motherboard has driver issues which failed to load during check, therefor producing the cold reboot.

Heating issues usually cold reboot also but i doubt you wouldnt have checked that as you mentioned.

So, since you are getting the cold reboot during "Games" aka during loading it highly indicates power issues.

You can start testing one by one rather easily.

PSU-->GPU-->Motherboard, test the Motherboard last cause chip issues that short dont always get reproduced therefor you might not figure out its that easily.

I only include GPU because PSU problems instantly reflect on the GPU since it is the only component that requires lots of Wattage during load and it will always trigger a bad PSU.

Unfortunately my job has nothing to do with computers haha. With the amount of stuff that always seems to go wrong with mine over the years it would definitely help! The RAM seemed to be the obvious issue given how it was fixed when I subbed in some different RAM (even though it then reappeared with the new set of RAM). Any thoughts as to why this is such an intermittent issue and not consistent over the month or two that in which it has appeared?

I'll have a suss of the PSU first. From a quick Google it seems like testing the PSU isn't easy at all! To be honest PSU's are the one thing I know nothing about. Some people are saying run Furmark, others are saying it needs to be physically tested using a multi meter et cetera. Is there something I can do to test it without actually testing it, say by simply upping the voltages via BIOS and then if the issue doesn't persist you can infer what the problem is? Obviously this would only work if the PSU is defective in a certain way I guess...

I'll test the GPU using Furmark/3DMark and see what comes up as well.

Thanks for all the replies!

[potis]

I obviously meant the "techshop" and not you ;p

Thing is, you dont test the PSU like that, multi meter works yes, for the people thats its their job to do so, for you its not.

You take it out, you plug another, you go a few days with a new one, or generally you try to trigger the problem with the new one, aka borrow one from a friend.

Problem persists-->Test next hardware.

[Blackmist]

PSU diagnosis often comes down to "plug in a different PSU, did the problem disappear?"

We've got multi-meters and all sorts of shit. Doesn't help unless parts are completely broken. How it acts under different kinds of load is the important bit.

I've had a machine that would switch off when a certain folder was accessed by another machine on the network. You could remotely browse other folders on the machine, you could browse the suspect folder on the local machine, but any attempt to look at it remotely caused a power failure. It was only when I swapped the PSU (it was a replacement I borrowed while waiting for mine to be repaired) that the issue went away. To this day, it's my "but that makes no sense" case.

[Goregrim]

Yeah definitely a frustrating experience! Just ran Furmark then, the second I started the test (1920x1080) the computer did the cold reset. Re-did it at a lower res and it lasted about 3 seconds before cold resetting/rebooting as per last time!

I seem to remember this computer technician saying that the PSU could be disregarded as he tested it, but it seems more and more likely that it's the issue...

[Remilia]

Yeah, like mentioned before, RAM stick problem would cause a BSOD, not a reboot.

Testing the PSU for a normal consumer is pretty much stick it into another machine and see if the thing happens to that computer at the same load. If that's not possible sticking a new PSU into your current machine. Otherwise find load testers, meters, and poke around inside, which I do not recommend you do. If it's a decent PSU and something is wrong with it, it may trigger one of varying protections it has equipped. Whether it's over temperature, over current, over/under volt, etc. Some motherboards have the option where it reboots immediately after a shut down for whatever reason. Should be able to turn it off in the BIOS (and I completely forgot the name of it).

Either way, it more than likely is the PSU though there is always the possibility the motherboard is at fault though unlikely (and bigger pain to test )

[Goregrim]

Yeah, like mentioned before, RAM stick problem would cause a BSOD, not a reboot.

Testing the PSU for a normal consumer is pretty much stick it into another machine and see if the thing happens to that computer at the same load. If that's not possible sticking a new PSU into your current machine. Otherwise find load testers, meters, and poke around inside, which I do not recommend you do. If it's a decent PSU and something is wrong with it, it may trigger one of varying protections it has equipped. Whether it's over temperature, over current, over/under volt, etc. Some motherboards have the option where it reboots immediately after a shut down for whatever reason. Should be able to turn it off in the BIOS (and I completely forgot the name of it).

Either way, it more than likely is the PSU though there is always the possibility the motherboard is at fault though unlikely (and bigger pain to test )

Ok yeah I'll definitely have to take it into the computer shop then haha.

It's strange, I've been running Furmark on and off for the latest few hours in 10 minute intervals, and at first like I wrote earlier it would crash instantly, now however I have no trouble running it at all (with both the burn-in selected and deselected and at the highest settings). Very strange.

I just don't understand how sometimes it appears that my PSU trips up and other times it seems content to work perfectly. The one time it did crash in this latest batch of tests was after increasing the fan speed settings in the BIOS (I ran the Furmark test immediately on startup and it crashed).

Also, with Furmark testing, under load the temp for the GPU does not go above 80 degrees celsius, which I'm assuming is fine?