Whea_Uncorrectable_Error BSOD keeps happening.

SilentHX

Reputable
Dec 24, 2015
8
0
4,510
It has been roughly 1 month before I completed my build. Ever since I installed Windows, I have been getting BSOD WHEA_UNCORRECTABLE_ERROR when there are moderate-heavy loads on my computer. I have tried many things, such as resetting windows and reinstalling all of my drivers; pretty much what the internet said about this error. I went back and realized the CPU fan wasn't sitting all the way down and so I just pressed down on it and I was able to reinstall Windows 10, which I had issues with reinstalling due to the same error mid install.

I'd really appreciate all the help that I could get.

Here are my parts:

CPU: Intel Core i3-6100 3.7GHz Dual-Core Processor
Mobo: Gigabyte GA-H110M-A Micro ATX LGA1151 Motherboard
Mem: Panram Ninja-V 8GB (2 x 4 GB) DDR4-2400 Memory
SSD: A-Data Premier Pro SP900 128GB 2.5" Solid State Drive
HDD: Seagate Barracuda 1TB 3.5" 7200 RPM Internal Hard Drive
GPU: XFX Radeon R9 380 4GB Double Dissipation Video Card
PSU: EVGA 500W 80+ Bronze Certified ATX Power Supply
 
Dec 5, 2014
147
0
4,710


Coincidental. The error message occurs with "A fatal hardware error". You also say it only happens when your computer is on moderate/high load. Yoou ALSO said you could get further with the install when you press down on your CPU fan.

I can guess that your CPU is overheating. I don't know much about cooling for desktops, but make sure you have applied enough thermal paste, and the fan has good contact with the heatsink or whatever its supposed to be.

Do you hear your CPU fan getting louder?
 

SilentHX

Reputable
Dec 24, 2015
8
0
4,510
I'll have to wait until tomorrow for thermal paste, I just settled for the pre-applied thermal paste; I've been told that that may be the problem. Thank you I'll update tomorrow to see if that is the problem.
 
Dec 5, 2014
147
0
4,710


I very much belive it is the problem
 

SilentHX

Reputable
Dec 24, 2015
8
0
4,510
I just finished reapplying the thermal paste and securing the fan into place, got the errors once again... I tried resetting the OS but it crashed mid reset due to the same error. I'm getting really tired of this and it's really stressing me out, is there any other way to resolve this issue?

Also the load required to crash it seems to have been brought down a lot, considering that I crash right after I login.
 

SilentHX

Reputable
Dec 24, 2015
8
0
4,510


Well the CPU temps seem fine but the clean install hasn't gone through. Tried installing Windows 7 but since my motherboard needs a patch and I don't have a Windows 7 disk, I'm out of luck. I've been focusing on school work, so I apologize on the late reply; haven't touched the computer since then and now I'm considering taking it to a PC shop even though I'll barely be able to pay for it.
 
normally a bugcheck 0x124 is called directly by the CPU by a error condition inside the CPU.
the memory dump contains a record of the reason the CPU shutdown the system. Often it is caused by two problems working together. A GPU using too much power from the PCI bus can cause the motherboard to reset the CPU, if you have a lower quality power supply, it is should keep the CPU reset until the power is stable again. it is common for them not to do this and the CPU just restart but the power is not stable so it gets a cache memory errorwhich the CPU internal cache memory controler detects, tells the CPU and the CPU shuts down with a bugcheck 0x124. This happens very fast (a few seconds)
the memory dump, has a system up timer, if this is the case the timer will show that the system was up for less than 15 seconds. If it is over 15 mins generally the CPU is overheating.

for the reset problem, you want to make sure your GPU is not overclocked and that the PSU is providing proper power. Also check any supplemental power connections to the GPU. if one is not connected and you run a 3D game or use any graphics accelerated browser the machine will reboot.
 

SilentHX

Reputable
Dec 24, 2015
8
0
4,510


According to Newegg the GPU I bought is pre-overclocked, could that be the issue?
 
Dec 5, 2014
147
0
4,710



COuld very well be, yes
 
yes, a pre overclocked GPU can pull a lot more power than the same GPU at stock speeds for the reference design.
you would want to make sure your PSU is within the recommended range for the overclocked GPU.
If your PSU is at the minimum specs or of a lower quality rating you might find that you have to reduce the overclock in the GPU back to the reference clock speeds.

But do check all the connections to the GPU here are the specs listed for your actual card:
Minimum 500w
recommended 650w <--- see page 3 below.
http://xfxforce.com/en-us/products/amd-radeon-r9-300-series/amd-radeon-r9-380-4gb-double-dissipation-r9-380p-4255

Sometimes if you have a silver or gold rated supply you can get away with the minimum.
also, if this card has two 4 pin connectors, make sure they are both connected to the power supply and come from different connections rather than a split or adapter. and make sure the lines are good. IE it is common for the one connector to get pushed out of the socket and not make a connection.



 

SilentHX

Reputable
Dec 24, 2015
8
0
4,510


How could I go about fixing it?
 
oh, I don't recall if I looked at a memory dump but make sure you are not running any GPU overclocking software. It is a bad idea to software overclock a GPU that is already hardware overclocked.
-------------
Well, provide the actual model number of the evga 500 w power supply
Make sure both PSU leads to the GPU are providing power.
Inspect each connections, test the PSU if you can.

I would then use a tool to underclock the GPU back to the reference clock speeds (or slower) and try the system out to see if it bugchecks.

from here: https://www.techpowerup.com/gpudb/2734/radeon-r9-380.html
it looks like the standard speed is 970MHz GPU clock and Memory Clock: 1375 MHz
5500 MHz effective with 2 GB RAM

your card has 990 MHz GPU core clock 1425 MHz memory clock 4096 MB RAM

The higher speeds and the double amount of RAM will mean the card will take more than the minimum 500 watt power when you actually use the 3d functions of the GPU.

Otherwise, I would put in a higher rated power supply, Your 500W supply has over current protection it has a control line that goes to the motherboard, power ok signal, if it detects too much current being drawn, it tells the motherboard and the motherboard logic tells the CPU to reset. (cheap power supplies don't do this and I have seen them actually just catch fire, same with cheap motherboards, I have seen graphic slots actually melt and catch fire.




 

SilentHX

Reputable
Dec 24, 2015
8
0
4,510


So the problem is that I got a PSU that doesn't support my build?
Here are the links: http://pcpartpicker.com/p/M4LNt6
 
the testing and review indicates that testing this power supply :"12v rail saw a substantial 0.38v drop."
http://www.hardocp.com/article/2013/12/16/evga_500b_500w_power_supply_review/4#.VpKW2hUrJjU

not very good for a overclocked video card with extra RAM.

I would also make sure your power supply is plugged directly to the wall outlet and not in a power strip.
If it is I would most likely try a upgraded power supply.