Computer Desktop shutdowns during Gameplay

[mapleteal]

For the past hour, my desktop has been shutting down without warning when I was playing a game, such as Dead by Daylight, Half-Life 2, or Rainbow Six: Siege.
The problem is, I thought I already took care of it. After I switched out my Lepa N500 PSU for an EVGA Supernova 550 G3 a few days ago, it was working fine until this. Running the GPU stress test via FurMark leads to a shutdown as well.

The CPU's idle temps are around 40s while being under load has 60s-70s.

CPU: i7-870, default
GPU: Gigabyte GeForce GTX 1060 WINDFORCE OC 6G, default
RAM: 16 GB DDR3 (2 x 8GB) Crucial
HDD: WD Green 1TB
MOBO: Gateway DX4840, BIOS update from American Megatrends is as of 04/01/2010

I'm really starting to hate this situation and I could use some help here.

 

[A_Goat]

Have you ever checked the 100% load temperatures on that CPU? Chances are they may be unfavorable.
Assuming they are high; Have you ever replaced the thermal paste on that CPU and cleaned the cooler? Could be overheating during load spikes and shutting down.

I had a similar OEM machine for quite some time (Same CPU), eventually it was just the motherboard that went out. They aren't the greatest especially for longevity.

 

[mapleteal]

Have you ever checked the 100% load temperatures on that CPU? Chances are they may be unfavorable.
Assuming they are high; Have you ever replaced the thermal paste on that CPU and cleaned the cooler? Could be overheating during load spikes and shutting down.

I had a similar OEM machine for quite some time (Same CPU), eventually it was just the motherboard that went out. They aren't the greatest especially for longevity.

I just did a five minute stress time through Intel's Extreme Tuning Utility and the max temps were 73, 74, 69, and 76 Celsius. I got these from Core Temp.
As for the thermal paste, I did replace it with a new one from Phanteks and cleaned off the old paste from the heatsink as well.

 

[photonboy]

1) not likely the CPU
2) not likely the PSU since you just changed it to address the same issue
3) possibly the motherboard, graphics card, or DDR3 memory

*First test to run is MEMTEST86.
1) download and create USB stick (or CD/DVD): https://www.memtest86.com/
2) change BIOS boot order as needed to boot then touch nothing and let it run
3) run until a FULL PASS or errors (very roughly 30min per 8GB)

The next and most likely suspect is the GRAPHICS CARD. I don't think that CPU has an iGPU so you'd need to swap in another card.

*alternatively try dropping both the VRAM and GPU frequency using whatever Gigabyte has for that. It's probably based on OFFSETS so just drop those down and if it works experiment by raising them or just RMA the card at that point since you'd know it's the problem.

 

[photonboy]

Update: since the DESKTOP alone apparently can shut down your system just disregard my comment about frequency. That's really to limit power draw which would not be high anyway for basic desktop.

I'd say it's likely the card but I don't know how to test unless you put it in another computer or try a different one in yours.

 

[A_Goat]

Have you ever checked the 100% load temperatures on that CPU? Chances are they may be unfavorable.
Assuming they are high; Have you ever replaced the thermal paste on that CPU and cleaned the cooler? Could be overheating during load spikes and shutting down.

I had a similar OEM machine for quite some time (Same CPU), eventually it was just the motherboard that went out. They aren't the greatest especially for longevity.

I just did a five minute stress time through Intel's Extreme Tuning Utility and the max temps were 73, 74, 69, and 76 Celsius. I got these from Core Temp.
As for the thermal paste, I did replace it with a new one from Phanteks and cleaned off the old paste from the heatsink as well.

Yes, now with those temps we know it's not the CPU temp causing a shutdown.
There aren't many cases where a GPU would cause a system shutdown unless it was in-fact power related, which I doubt. However if the system was locking up instead of shutting down, that would be a red flag for the GPU.

 

[mapleteal]

1) not likely the CPU
2) not likely the PSU since you just changed it to address the same issue
3) possibly the motherboard, graphics card, or DDR3 memory

*First test to run is MEMTEST86.
1) download and create USB stick (or CD/DVD): https://www.memtest86.com/
2) change BIOS boot order as needed to boot then touch nothing and let it run
3) run until a FULL PASS or errors (very roughly 30min per 8GB)

The next and most likely suspect is the GRAPHICS CARD. I don't think that CPU has an iGPU so you'd need to swap in another card.

*alternatively try dropping both the VRAM and GPU frequency using whatever Gigabyte has for that. It's probably based on OFFSETS so just drop those down and if it works experiment by raising them or just RMA the card at that point since you'd know it's the problem.

I ran MEMTEST86 overnight and it had two passes and no errors for 7 hrs and 23 minutes.

 

[mapleteal]

Turns out I need to take out the GPU, blow out any dust on its connector to the MOBO, and secure it in the slot as tightly as possible and screw it in tightly too. I also have the PSU plugged into the wall, not on the power strip. I'll leave this thread open for any more shutdowns.