Random shutdown and restart while playing games

[Iggy Wolf]

So recently replaced my PSU because its bearings and fan were failing, and I assumed the shutdown and restart during heavy gaming was caused by it. However, even after replacing it yesterday, after playing Gears of War 4 for several hours, my PC randomly shut down and restarted again. It should be noted that it was constantly on for the past 10 years before I started playing due to downloading the game.

Prior to that, it shut down while playing AC4 Black Flag. My GPU temps were showing as 76 C and my CPU temps were 68 C. So it's not like it was overheating. So what could be the problem? Once again, the PSU is brand new and normally I can play for a long time without this occuring. It should be noted it DOES NOT occur outside of gaming (hence, it managed to run for 10 years downloading Gears of War 4 without a problem).

My system specs btw are as follows:
Intel Core i5 6400 Skylake at 2.7 GHz
ASUS STRIX Nvidia Geforce GTX 970 OC 4GB VRAM
Windows 10 Home Edition
PSU: Thermaltake TR2 600W

[Cathulhu]

10 years to download Gears of War 4? Ouch. Must be a really slow connection.

Anyhow, are there any interesting entries in the Windows event log at the time of the restart?
Does it really shut down, or just restart without showing a bluescreen?
Shut down means turning completely off requiring you to press the power button to start it.

[Iggy Wolf]

Opprinnelig skrevet av Cathulhu:

10 years to download Gears of War 4? Ouch. Must be a really slow connection.

Anyhow, are there any interesting entries in the Windows event log at the time of the restart?
Does it really shut down, or just restart without showing a bluescreen?
Shut down means turning completely off requiring you to press the power button to start it.

That was a typo. I meant to put "10 hours". And no, it simply shows a "Critical: Kernel/System Error ID 41. The system may have shutdown unexpectadly etc etc.". It shut down, and then restarted without a BSOD, but only because my startup and recovery settings had "Restart computer after shut down" under the "Power Failure" settings.

[Cathulhu]

Could you disable that setting so it would show you an actual bluescreen instead and note everything in case it happens again?

[hawkeye]

My guess - the ripple control on the psu's 12V rail can't handle the overclock on the gtx 970.

[Iggy Wolf]

The thing is though, the PSU I bought is the SAME one (brand and type) as the old one, and the old one managed to run for 5 years for me without this problem. I should mention though that I played Gears of War 4 for several hours AFTER the computer had been running non-stop for 10 hours downloading the game, if it matters.

Edit: I just noticed my Boost Clock and Memory Clock seem to have fluctations (jumps) from minimum to maximum. It doesn't occur continously or consecutively but randomly (although I was just messing with the Memory Clock in the GPU Tweak settings). I'm also simply browing on my PC, not doing any gaming right now.

[hawkeye]

You bought the same sticker as the old one, not necessarily the same psu. Thermaltake is just the retail name, the manufacturer is someone else and could change over time.

Thermaltakes and gtx970s were notorious for 12V issues. I had one, bought a Corsair, problem solved.

If you want to test it, use msi afterburner and underclock the gtx 970.

edit - if you have been messing with the oc, you could be the cause. Run msi afterburner. Press the reset button.

[Iggy Wolf]

Opprinnelig skrevet av hawkeye:

You bought the same sticker as the old one, not necessarily the same psu. Thermaltake is just the retail name, the manufacturer is someone else and could change over time.

Thermaltakes and gtx970s were notorious for 12V issues. I had one, bought a Corsair, problem solved.

If you want to test it, use msi afterburner and underclock the gtx 970.

edit - if you have been messing with the oc, you could be the cause. Run msi afterburner. Press the reset button.

I wasn't messing with it, since I didn't actually change the settings. I was just simply testing to see if I could, but I was doing this AFTER the shutdown, not before. At the moment, what could I do? Can I underclock my GPU and by how much in order to compensate for the 12V?

[hawkeye]

Msi-ab has 3 main sliders - power limit, core and memory clocks. When started it will show what the gpu is set to. If it is oc'ed the figures will be +ve. Pressing reset sets everything back to normal settings. Try that. If it still crashes, underclock core and memory by 200 each. Try again.

[Iggy Wolf]

Opprinnelig skrevet av hawkeye:

Msi-ab has 3 main sliders - power limit, core and memory clocks. When started it will show what the gpu is set to. If it is oc'ed the figures will be +ve. Pressing reset sets everything back to normal settings. Try that. If it still crashes, underclock core and memory by 200 each. Try again.

Mine's not MSI but ASUS STRIX. It already comes with a GPU Tweak, allowing me "offset" each setting (Memory Clock and Boost Clock) by +10 or -10 while scrolling the mouse wheel). I can offset the Memory Clock and Boostclock by -200. How will this affect my in-game performance though?

[Iggy Wolf]

Just curious, can the problem be the GPU itself? It seems like even if my PC is cool, it can shut down just after two hours of gaming. For the record, I was playing Gears of War 4 and it was using up to 4 GB VRAM, even though the card only has 4 GB VRAM max. Also, before shut down, the GPU MV Voltage dropped to 40.0.

[havelcøde]

As stupid as it might sound... have you checked that the motherboard power connector is properly seated?

I had a similar problem where the connector wasn't fully seated so as soon as my consumption increased when I started playing, my PC would reboot because it couldn't supply enough power.

[Fluffy]

you said it was restarting even before the new psu dont get me wrong thermaltake tr2's are not the greatest power supplies but the 600w variant shouldnt have an issue with your system as even with the gtx 970, its fairly low wattage

i would suspect overheating of cpu, nvidia drivers (did u try updating them/clean install) or memory errors (run windows memory diagnostic or memtest). ive seen a good numbber of pc's crash during gaming due to faulty memory as the game uses high gb/taxes the memory alot more than regular programs

[Iggy Wolf]

Opprinnelig skrevet av Fluffy:

you said it was restarting even before the new psu dont get me wrong thermaltake tr2's are not the greatest power supplies but the 600w variant shouldnt have an issue with your system as even with the gtx 970, its fairly low wattage

i would suspect overheating of cpu, nvidia drivers (did u try updating them/clean install) or memory errors (run windows memory diagnostic or memtest). ive seen a good numbber of pc's crash during gaming due to faulty memory as the game uses high gb/taxes the memory alot more than regular programs

Well, the CPU was showing temps for cores only in the 65-75 range. If it matters, it did this during Gears of War 4, with my settings maxed out and the card utilizing ALL 4 GB of VRAM. It put my memory usage at 3840.0. My GPU temps were at a consistent 75. So I don't think my CPU was overheating.

I mean, it also shut down under AC 4 Black Flag, and that game isn't even considered demanding by today's standards (minus the poor optimization). So far, this is the ONLY 2 games it has done this with. I haven't tested it with other games. Someone above did mention that though the wattage is high, the ampage on the 12V line could be too low, especially if the higher demanding games require more power consumption.

Up until now, most of my games have been not been demanding like any of the newest generation (the newest ones I have so far are Hitman, COD MWR, and COD WWII. And even with COD WWII's 100% CPU usage, it didn't shut down after playing for several hours. So I don't know if the thermal paste dried up due to that or the GPU's power consumption requirements magically went up. Wouldn't faulty RAM sectors cause other problems? I mean, my computer without gaming is able to work just fine. Gears of War 4 was only using 50% of the system RAM (I have 16 GB RAM). COD WWII was using 70%.

[Fluffy]

Opprinnelig skrevet av Iggy Wolf:

Opprinnelig skrevet av Fluffy:

you said it was restarting even before the new psu dont get me wrong thermaltake tr2's are not the greatest power supplies but the 600w variant shouldnt have an issue with your system as even with the gtx 970, its fairly low wattage

i would suspect overheating of cpu, nvidia drivers (did u try updating them/clean install) or memory errors (run windows memory diagnostic or memtest). ive seen a good numbber of pc's crash during gaming due to faulty memory as the game uses high gb/taxes the memory alot more than regular programs

Well, the CPU was showing temps for cores only in the 65-75 range. If it matters, it did this during Gears of War 4, with my settings maxed out and the card utilizing ALL 4 GB of VRAM. It put my memory usage at 3840.0. My GPU temps were at a consistent 75. So I don't think my CPU was overheating.

I mean, it also did this under AC 4 Black Flag, and that game isn't even considered demanding by today's standards (minus the poor optimization). So far, this is the ONLY 2 games it has done this with. I haven't tested it with other games. Someone above did mention that though the wattage is high, the ampage on the 12V line could be too low, especially if the higher demanding games require more power consumption.

Up until now, most of my games have been not been demanding like any of the newest generation (the newest ones I have so far are Hitman, COD MWR, and COD WWII. And even with COD WWII's 100% CPU usage, it didn't shut down after playing for several hours. So I don't know if the thermal paste dried up due to that or the GPU's power consumption requirements magically went up. Wouldn't faulty RAM sectors cause other problems? I mean, my computer without gaming is able to work just fine. Gears of War 4 was only using 50% of the system RAM (I have 16 GB RAM). COD WWII was using 70%.

did you test the ram?? idk why your telling me all the reasons it couldnt be ram when a simple test would rule it out. your free to just assume its the new psu and not test other components and buy another one..

just because it only uses 50% of the ram doesnt mean it couldnt be encountering faults in the ram.. your cpu temp is fine i highly suggest testing the ram

[Iggy Wolf]

Opprinnelig skrevet av Fluffy:

Opprinnelig skrevet av Iggy Wolf:

Well, the CPU was showing temps for cores only in the 65-75 range. If it matters, it did this during Gears of War 4, with my settings maxed out and the card utilizing ALL 4 GB of VRAM. It put my memory usage at 3840.0. My GPU temps were at a consistent 75. So I don't think my CPU was overheating.

I mean, it also did this under AC 4 Black Flag, and that game isn't even considered demanding by today's standards (minus the poor optimization). So far, this is the ONLY 2 games it has done this with. I haven't tested it with other games. Someone above did mention that though the wattage is high, the ampage on the 12V line could be too low, especially if the higher demanding games require more power consumption.

Up until now, most of my games have been not been demanding like any of the newest generation (the newest ones I have so far are Hitman, COD MWR, and COD WWII. And even with COD WWII's 100% CPU usage, it didn't shut down after playing for several hours. So I don't know if the thermal paste dried up due to that or the GPU's power consumption requirements magically went up. Wouldn't faulty RAM sectors cause other problems? I mean, my computer without gaming is able to work just fine. Gears of War 4 was only using 50% of the system RAM (I have 16 GB RAM). COD WWII was using 70%.

did you test the ram?? idk why your telling me all the reasons it couldnt be ram when a simple test would rule it out. your free to just assume its the new psu and not test other components and buy another one..

just because it only uses 50% of the ram doesnt mean it couldnt be encountering faults in the ram.. your cpu temp is fine i highly suggest testing the ram

Cause I imagine faulty RAM would cause problems in other aspects of my PC. It only does this during high demanding games. It doesn't do it while the PC is idle and I'm browsing and/or playing regular games.

It also doesn't shut down immediately but only after prolonged hours of play. In fact, the 2nd 8 GB stick I had only recently added. I'm not saying it's a problem to test the RAM. I'll probably do that if the new PSU doesn't solve the problem.