Some advice to more rapidly identify failing hardware overclocking.
Some advice to more rapidly identify failing hardware overclocking.
I notice my build details appear in my profile.
Right now I'm testing RAM configurations and cache overclocking for fun (already having a stable CPU overclock).
Recently, my PC has become quite unstable.
Initially, I assumed the issue might be due to incorrect RAM settings, after pushing it to 2400MHz with specific voltages and System Agent settings.
But now it crashes at 2400MHz with 13-16-16-28 and 2T 1.42V, which seems unexpected.
This makes me question whether the problem lies in the RAM or if the CPU cache is affected.
I recently increased the cache to 4.2GHz while keeping the voltage constant, based on past stable overclock experiences.
However, during the latest crash, I reduced it to 4.1GHz.
The error messages from the motherboard usually point to RAM issues, but I’m not sure.
I’m considering reverting to default settings and checking stability over a week before tackling the cache or CPU again.
Some people suggest observing whether a crash indicates CPU or RAM failure—either freezing first or stopping abruptly.
TL;DR: Can I identify hardware overclock failures based on crash behavior? (Please read the post)
EDIT - MY BUILD:
Asus X99-A with BIOS 2101 | Intel i7 5820k OC @4.4GHz 1.264V and Noctua NH-U14S |
2x Corsair Vengeance DDR4 8GB at 2400MHz 10-12-12-28 1T 1.4V | Asus GeForce GTX970
STRIX OC | EVGA 750 G2 750W | Samsung 850 Evo 1TB | Windows 10 64-bit |
Be-Quiet Silent Base 800 w/ Silent Wings | 2x Dell U2414H OC @72Hz with Display Port
Port (my signature not visible?)
Well, no one had explained anything clearly yet, so I decided to just put RAM to stock and see how it goes. If you're still having the same problem, make sure your clock is disabled in the cache. Generally, when the cache is unstable, your screen will freeze and RAM issues often cause BSODs, sometimes with a warning in Event Viewer. Windows RMB>event viewer>windows log>system works best for me.
Typically, but not always? It seems mostly focused on identifying possible causes of crashes.
What voltages should be applied to the cache? Are they comparable to the core voltages?
Didn't clarify how your system is unstable. Are you experiencing BSOD, freezing, lag? During stress tests or regular use? Gaming? What led you to believe your system is unstable?
I can't pinpoint the exact issue from across the globe, but I'm here to assist. This is why I wanted to understand the details you mentioned.
Make sure you've already checked other forums too—overclockers.com might be helpful.
My mistake, here you are.
It mostly freezes for 3-4 seconds before showing a BSOD with the CLOCK_WATCHDOG_TIMEOUT error. At times I can't escape the crash and have to reset the motherboard.
This occurs only in games that use real-time rendering, unlike many FPS titles that render when loading from the menu.
I usually play games like Ark or Just Cause 3 for about 1.5 hours before a crash might happen (though not always). In Minecraft, I try to load a saved world right away—if it doesn’t crash there, it shouldn’t crash during a session.
Despite running a full system stability test in Aida64 for over two hours, this issue still exists.
I’ve adjusted the RAM to XMP settings and increased the cache again to check if it triggers another crash (unstable cache would be the sign).
For now, I’m keeping this thread open to track progress; I might share updates on overclockers.com or elsewhere if necessary (I try to keep it private since comments can be harsh).