Need help pinning cause of "internal parity error"

Colonel Sanders

Supreme [H]ardness
Joined
Sep 26, 2001
Messages
5,592
Alright, so here's the background. I have been running the CPU/mobo in my sig since 2012. When I first built this rig I had it overclocked to 4.8ghz but I eventually ran into some stability issues after a few weeks, so I turned it down to 4.6ghz and never had a problem since. This system has been rock solid stable - IntelBurnTest, Prime, gaming, whatever, I've never once had a stability issue.

Fast forward to the past couple weeks. Suddenly I'm getting hard freezes with looping sound when gaming (Mad Max, MGS5, etc.) I might be able to game for 2 hours or 10 minutes before it happens, but it usually does happen.

The only evidence/error message I can find is a WHEA-Logger error in event viewer that says:

A corrected hardware error has occurred.

Reported by component: Processor Core
Error Source: Corrected Machine Check
Error Type: Internal parity error
Processor ID: 6

Some google-fu found a few threads on other forums about the exact same processor (3570k) when overclocked and some people mentioned upping voltage. Tried that, no difference (at least to the level I tried.) I didn't want to up it much more than initial testing because why would I all of a sudden need a huge boost in voltage when nothing else in my system changed? I also tried setting Load Line Calibration in my Sabertooth Z77 to Extreme (was on Very High) with no difference.

The hard locks seem limited to gaming, but I did test with IBT on Extreme. After 10 minutes or so I did pick up another "corrected parity error" in event viewer (this was before I messed with voltage or LLC) and after tweaking voltage/LLC, I was able to run 100 passes of IBT on Extreme with zero issues or events.

Temps are just fine, absolute single core maximum was 87c after 30 minutes of IBT. Otherwise during gaming it's in the lower 70s or even 60s. GPU temps are about 80c for the most taxing game. So yeah, temps are more than good.

Anyone have any ideas? Does this particular event guarantee CPU-related issue or could it be GPU? The 980 Ti was the most recent change in my system (with the Samsung SSDs next most recent but several months ago.) Issue started weeks after installing the 980 Ti, however.
 
are you running your system RAM at anything different from stock?. generally I found those kind of issues related to faulty RAM. try lowering RAM speed and do some testing you can start from 1600mhz (the recommended spec for ivy bridge) and keep increasing the frequency until you find no error even if that mean again 2400mhz just to discard any issue with RAM or memory controller. what voltage are you using for those 4.6ghz?
 
are you running your system RAM at anything different from stock?. generally I found those kind of issues related to faulty RAM. try lowering RAM speed and do some testing you can start from 1600mhz (the recommended spec for ivy bridge) and keep increasing the frequency until you find no error even if that mean again 2400mhz just to discard any issue with RAM or memory controller. what voltage are you using for those 4.6ghz?

Alright, I'll check RAM next. Right now I removed the OC on the CPU, running at stock settings but kept the RAM at 2400mhz (the default XMP profile/timings/voltage.) That's the way I've been running the RAM since I installed it probably 2 years ago now.

I had the CPU at offset +0.030v. Under load that translated to around 1.252v
 
Alright, I'll check RAM next. Right now I removed the OC on the CPU, running at stock settings but kept the RAM at 2400mhz (the default XMP profile/timings/voltage.) That's the way I've been running the RAM since I installed it probably 2 years ago now.

I had the CPU at offset +0.030v. Under load that translated to around 1.252v

I think at that speed you should be able to toy a bit until 1.270v if the chip is really degraded.. but first test RAM.
 
Interesting, I occasionally have that freeze when I open up MSI afterburner, whocrashed might help you
 
Back
Top