WHEA-Logger - Internal Parity Errors?

hangar34

hangar34

hangar34

Associate
Joined
28 Jul 2013
Posts
17
Hi,

I have recently been seeing an increasing number of WHEA-Logger errors similar to this:

A corrected hardware error has occurred.

Reported by component: Processor Core
Error Source: Corrected Machine Check
Error Type: Internal parity error
Processor ID: 2

Could someone kindly give me some advice about how to see whether this is something fixable or whether permanent damage/deterioration of my system could be the cause. My PC is an Overclockers built machine of about 4 years old.

The thoughts I have had are:

a) Slightly increase VCore voltage in the BIOS to see if this stabilises the overclocking.

b) Remove the CPU and check the heat compund and renew.

Could this error relate to any other components that may be on their way out (e.g. PSU, GPU etc.)?

Thanks in advance for any help offered.
 
for a 4yr old machine i would remove all the dust build up from the heatsinks and fans, clean off the old thermal paste and replace with some fresh artic silver 5. if errors not cured then yes increase vcore a notch or 2.
id probably clean the gpu heatsink too if its never been done, while you're in the de-dusting mood :P

ive not had a psu deteriorate they tend to just go bang around 4-5yrs old :|
 
zipp0r, thanks for your reply. As regards dust, this shouldn't be a problem as I check the inside of the system regularly and it's surprisingly dust-free. However, thanks for confirming that the thermal paste could probably do with renewing and the small adjustment to the vcore setting.

I'll post back with how I get on once I've done it.
 
Right, I've upped the VCore by a couple of small steps and completely stripped the CPU heatsink down this morning, reapplied thermal paste and reassembled. Re-seated the RAM and blown what little dust was in there out. PC has booted normally and Windows Error Logging has been cleared and set up for me to view if there are any more WHEA-Logger errors.

Incidentally, how much can I increase the VCore voltage before I'm doing something inadvisable?
 
Sadly, another batch of 4 or 5 WHEA-Logger errors this morning. These are the last settings I was sent by OC for my system when I had to reinstate the overclocking...

Intel Core i5 3570K 3.40GHz @ 4.50GHz DDR3 Ivybridge:

CPU Clock Ratio - 45
System memory multiplier - 16.00
Performance Enhance - Standard,
DRAM Timing Selectable - Quick
Cas latency - 11
trcd - 11
trp - 11
tras - 30
command rate - 2
PWM Phase Control - exm pref
Voltage response - Fast
Vcore Loadline Calibration - Turbo
CPU Vtt Loadline Calibration - Turbo
CPU Vcore - 1.33v
CPU Vtt - 1.15v
DRAM Voltage - 1.65V

Any advice on whether these are sufficient or might be causing the WHEA-Logger errors would be appreciated.
 
Last edited:
what is the vcore showing under full load on all cores in cpuz ? what is the cpu temp shown in coretemp under full load? what heatsink are you using ?
how many sticks of ram and what is the default specs, really 11-11-11-30-2 1.65v ??? cpuz should be able to read the memorys default timings and voltage.

what was the pc doing, i.e did the errors appear at full load 4.5 or at idle 1.6 ??
 
EDIT: How best should I put the CPU under load to read the vCore voltage? Where can the CPU temp be found?

I don't have CPU-Z and have no idea how to use it, but will download and install it. I can only therefore part-answer your questions...

I have no idea of the heatsink manufacturer, but it is a large cuboid shaped assembly of horizontal metal wafers, with copper pipes at the bottom and what look like copper 'nipples' at the top. There is a large, fan clipped to the side blowing air across the whole assembly. It's large enough to fill the depth of the case from CPU to side panel.

The RAM is 2x8GB (16GB) DIMMS of CORSAIR CML16GX3M2A 1600C10 1600MHZ 10-10-10-27 1.50v

The WHEA-Logger errors today were during a period of playing ARMA3 this morning, although I have seen them in Windows Error Logging following other CPU-based operations (e.g. 3D rendering). I have not had any BSOD's and would probably not have been aware of the WHEA-Logger errors if I hadn't been looking through the Windows Error logs.

I will run CPU-Z and post some other data when I have it.
 
Last edited:
While running CPUID's HWMonitor, I was able to generate a WHEA-Logger internal parity error by conducting a 3D render test...

These were the MAX temps recorded:

Core#0 70C
Core#1 72C
Core#2 73C
Core#3 75C
Package 76C

Max VID 1.221 V
 
Last edited:
Sorry, my bad... I didn't realise the vCore voltage was under the mobo data block.

Whilst I haven't been able to force another WHEA-Logger error, the MAX voltage recorded by HWMonitor was 1.368v. The current vCore voltage is set to 1.344v. Therefore, I need to increase the value in the BIOS right? To what setting... 1.368v or slightly higher?

Thanks
 
no dont increase it anymore, its getting abit too much now.
maybe try dropping the cpu multiplier until these errors stop. try to lower the vcore too.
 
EDIT: I have adjusted the vCore back down to 1.33v and changed the CPU ratio (multiplier?) to 43 (4.3GHZ)

Thanks zipp0r for the ongoing help. I'll try a small step down on each and see what happens. Any recommended values you would suggest?
 
Last edited:
well you might be able to go as low as 4.3ghz with say 1.225v, 4.4ghz with 1.250v or similar to that. maybe read some 3570k overclocking threads to see what other people were getting? pretty sure they never needed to go so high with the vcore.
 
You must log in or register to reply here.
Back
Top Bottom