System reboots when gaming - Help needed

[Roblob2]

Hi,

I need some help with the problem of a Win10 machine starting to reboot while gaming.

I have a about 5 year old build (below) on which I've been upgrading components at times as needed. Most recent changes have been more memory (up to 16GB) about 5 months ago, a new graphics card (ASUS GTX 970 Strix) about 2 months ago and of course the upgrade from Win7 to Win10 in July. This setup has been running fine up until recently.

Problems started about 1 week ago when the machine started randomly rebooting when under load (playing games, Sims 4 seems to be a surefire way to reboot). There doesn't seem to be any consistency as to when the machine restarts except that it now does so almost always at some point during gaming. The restart might be 2 hours into the session or it might happen in 10 minutes.

As a first step I removed my graphics drivers (with Display Driver Uninstaller) and reinstalled an older nVidia driver (353_62) since the newer version was updated on 2016-09-20 and I figured it might be causing problems. No help from this.

I then took the machine apart and cleaned everything, I also changed the thermal paste on the CPU and refitted the cooler. I've setup monitoring (I've tried CAM, CPUID HWMonitor & now SpeedFan for logging) and to me the CPU seems to be running quite hot. It idles at around 50C (no ambient temp load) and shoots to 70-80C under load.

I ran IntelBurnTestV2 for a little while and the CPU temp stayed at 80C which to me seems hot although it didn't show any instability. The games seem to be the only ones to cause the restarts and from what I've seen it isnt' necessarily CPU temp related as the monitors have shown under 70C temps when the system suddenly restarts.

Memory tests didn't find any problems with the system memory so I ran FurMark but it didn't seem to do much for the GPU: temp leveled at 65C, power at 80% and fan didn't rise above 35%.

Next step was running OCCT on automatic and it ran for about 3.5 minutes until CPUTIN hit 85C and it stopped. Other values seemed ok, voltages didn't seem to tremble.

Then I fired up Sims 4 and loaded up a save game. In less than 5 minutes the machine rebooted without warning. SpeedFan logged CPU temp at 72C and GPU at 61C at the reboot. The only other change from the logs was +12V which dropped from 11,51 to 11,46 a little before the crash, don't know if this is a significant change. I've tried getting Windows to go to BSOD instead of a restart but that doesn't seem to happen, the machine just reboots.

Now I'm at a loss as to how to proceed?

Any help would be much appreciated.

BUILD:
Intel Core i5 750 (running stock frequency, stock cooler)
ASUS Maximus Formula III P55 Mobo
Kingston DDR3 4x4GB memory
ASUS GeForce GTX 970 Strix (running stock frequency)
Nexus RX-8500 850W PSU
WDC Caviar Black 500GB HD
Corsair Force 3 240GB SSD
ADATA SX900 240GB SSD
Kingston 240GB SSD

 

[GameFreak01048]

Hello!

It seems as though your specs are fine except your CPU, what might be happening is when you tell your PC to work and work hard (playing games for example) your GPU kicks in but your CPU falls over itself trying to keep up so it reboots.

Your CPU is very old so I suggest replacing your CPU with a newer CPU but this will also mean you will have to change your motherboard as well (most likely).

I hope this helps

 

[dankcik09]

power supply. these can be difficult to test. but its kind of one of those things that once you have tested everything else it leaves this red flag, and many people dont ever think to look at this. hard to diagnose, but common problem.

 

[Roblob2]

Just a quick update. I watched HWMonitor readings a bit more closely while playing and under my mobo in Powers it displays CPU consumption as 168W but Currents hit 168A. That can't be right, can it?

Under CPU it displays Package as 95W and my GPU was drawing about 65% so that would be about 95W (of 145W).

EDIT: I'm not sure if I'm calculating this correctly, but could I be overloading the 70A rail my cpu, gpu and mobo is drawing power from?

 

[Roblob2]

Okay, now I'm truly baffled. Since my CPU was pretty old and PSU might have been faulty too, I figured I'd build a new computer around my existing parts. I bought a new CPU (i5 4460) and mobo (Asus z97-P), a new PSU (Fractal Design Edison 750W) and a new case (Fractal Design R5).

Together these new components should fix any issue I had that was causing the reboots under load. I got the new machine together last night and today I was gaming when the system once again rebooted. WTF?

Windows events report Kernel power problem once again and from the logs the machine seems to have suffered one last night too at 03:00 when no one was using it.

What else can be causing this? Can it be faulty mains power or is there something of the existing components (GPU, memory, drives) that could be causing this? Can a faulty GPU, for instance, cause power failures that reset the system?

 

[Roblob2]

Apparently this has all the time been an issue with my Asus GTX 970. I'm not sure if my graphics card is faulty or if this is a driver issue with Win10, but replacing all the other components didn't do anything to fix the issue. Although now the symptoms are just the displays going blank (with various colors) instead of simply rebooting, which I guess is due to the new mobo handling issues better.

I've tried various driver versions and tweaks to my GPU, but so far only underclocking the GPU seems to bring any sort of relief if not a real fix. Reflecting on the circumstances when the problems started I'd wager that the Win10 Anniversary update seriously fouled up my system. Before that update my system was running rock solid.

So if you run into similar problems try underclocking your GPU and then pray for driver updates from NVidia.