BSOD when using CE!

[oronnoro]

Hello all,

I'm having an issue with my system. I think I've narrowed it down to the CPU, but I would like to know the opinion of others.

I had the system for a few months now with no issues, then about 2 weeks ago, the system started having issues. It would restart without any kind of warning. I thought it was because of the new windows update, I was running the latest Dec 2020, version. I did a system restore and went back to the 1906 Win10 version. Since then, I've only had 1 crash and that was the same day as the restore and that's it, it's been about 3 day or so since the restore.

Today, I wanted to play a game, but I wanted to be powerful, so I used CE to do that. When searching for stat values using CE, the computer would BSOD. I have not used CE for a while since I've build my new PC, the crashes started without the use of the program, I believe the use of CE has helped me narrow down the actual issue, but I'm not sure.

What's strange is that, the computer would only BSOD, if the clock ratio was left at, default, 49 and would only crash when searching "all" in value type. If I change my CPU clock ratio to 46 and use CE, everything works just fine, I've tested this multiple times now and going back to default settings causes a BSOD.

My question is, since it looks like the CPU has gone bad, has it actually gone bad and should I ask for a replacement? My conclusion is that, when searching for these values, it's accessing the CPUs memory, or something along those lines, and not getting what it wants... I'm not sure.

Clearly I'm no expert at computers or CE, so I'm not sure what's going on, I know that downclocking the CPU to only 4600MHz works and does not cause any BSOD, I just want to know why, what's CE accessing that would case a BSOD at default cpu settings...

I will most likely ask for a replacement, but I would like to know if anyone can provide any advice or other kind of information.

The few errors I've gotten:
fault in nonpage area - I searched this up and saw people suggesting, disabling paging, I tried that and it did nothing.

driver overran stack buffer

Intel - 10900k
Nvidia - ASUS 3080
Ram - 32GB
Windows 10


Thank you very much!

[predprey]

Unless you were testing with DBVM, chances of a BSOD due to CE is rather low, even then DBVM rarely crashes for me, and your error messages don't seem to look like those from DBVM BSODs.

So you're probably looking at an overclock that barely worked for 2 weeks until the CPU degraded just enough for it to become unstable, or you just didn't encounter heavy workloads enough to destabilize it. In this scenario, you can either up the voltage if the instability is due to power starvation, which would again reduce the CPU's lifetime but shouldn't matter if you expect to upgrade it again in 2-4 years. Or, you can just adjust the clock down till its stable again.

From the BSOD errors, seems like it might be due to bad RAM overclocks too. So run memtest and prime95 tests to check for stability and adjust your memory or CPU clocks accordingly, else call up your retailer to complain about the bad overclocks within 2 weeks and get a replacement on warranty, cause 49 does not seem like the stock clock ratio for a 10900k to me.

[oronnoro]

The thing is, I've had random restarts without any kind of warning or bsod. The crashes ware happening for about 2 weeks, then I did a system restore, 3 days ago, at which point things seemed to be working just fine at default settings.

Only when I used CE did my pc crash with the bsod. Changing the clock ratio from default 49 to 46 stopped the crashes. And the 10900k has a default ratio of 49. I wanted to be sure, so I looked after you made your post.

I've tried rearranging my ram sticks, even just trying one by one the pc crashed if it was using default clock rates.

I'm also not using DBVM, I've never used it. I'm a grunt at the lowest level when it comes to CE, I just don't know that much, but enough to at least get me some money or change some stats around.

I know it's a strange issue to ask here, but I've tried elsewhere and got told to move along because talking about cheating is a touchy subject. I appreciate all the help anyone here can chime in. Thank you.


Edit:
The one thing that's strange is the cpu clock rate... From what you've said, maybe the cpu has gone bad... I'll do a memtest later tonight to make sure, but like I said, I've removed all the ram and tried one by one with default settings, and as long as the cpu was at clock ratio 49 the pc crashed when searching "all" in value type.

[predprey]

Well, if the CPU still BSODed even with every single ram tested individually then it's most likely the CPU's fault. The 10900K has a base processor frequency of 3.7Ghz and a turbo frequency of 5.3Ghz, which translates to a CPU multiplier of 37 at baseline and 53 on turbo. Turbo by default only activates if the CPU detects that not all cores are being used and the max turbo frequency goes up to 5.3Ghz when only one core is being used by default.

But when you overclock in the UEFI/BIOS, you are setting all cores to run at 4.9Ghz when you set the multiplier to 49. Sorry to say, but you most likely lost the silicon draw badly and your CPU can't even overclock to 4.9Ghz stably, so you can either inch up the voltage little by little and see if it helps or RMA it if overclock stability is in your warranty clause.

But note that sooner or later the new stable voltage would become unstable again and you would have to inch it up again, hastening the demise for your processor, it's kind of like being addicted to drugs. Until one day either the voltage or high temps from the voltage kills your processor due to electromigration. But you'll probably be changing your CPU by then which is 2 to 4 years down. And if not, it is quite likely that you can still use your processor at stock clocks even when all overclocks are no longer stable, which can last you another 2 to 4 years or more.

Memtest and Prime95 are different tests so you should do both, memtest tests for faulty ram, prime95 tests for processor stability. just be sure you have proper cooling or some small ffts prime95 tests can generate large amounts of heat and might damage your processor if not cooled properly.

[oronnoro]

Heh. I've gotten more help form a forum that's about cheating then from 2 different teck sites. Thank you for going the 100km for me.

Sorry if I'm complicating things, I'm no expert when it comes to overclocking and the like. I've left things at default, yes, the mobo shows that the CPU is set to 3700MHz, but I was using the clock ratio 49 for some stupid reason to explain that, that's the speeds it goes at when on the desktop (I used HWMonitor to see what it goes to). The actual setting is set to "auto". I just wanted to say that, when I do change the clock ratio to something other then auto, like 46/47, doing the same CE test, value type "all", the PC does not crash.

Sorry for a terrible explanation on the previous posts.

Everything is on default, with the exception of that one feature, Enhanced Turbo, which was recommended to be turned off by a few tech channels on YT, like Gamers Nexus, for high voltage or whatever, and changing the ratio to 47 for the time being.

So, the memtest finished a little bit ago, came back with no errors. I've yet to do the prime95 test, I'll do that one later today, I'll do that on default, so it'll be at 4900MHz.

Again, thanks for everything!

Edit:
Does CE look at the gpu memory, or is there a test I can do for gpu memory as well? I'll google this one and see if I can find anything.

Edit2:
I've defaulted everything once again, and CE is not crashing my PC when I run the scan again... Well, I probably just jinxed it... Not sure what's going on anymore.

Thanks again for all the help. :

[predprey]

@Edit:
CE does not use GPU memory, and the chance of incompatibilities between CE and GPU here is extremely low. You can remove every unnecessary cards
and USB devices to narrow down the troubleshooting, though it probably would not yield much insights or benefits and is more trouble than worth.

@Edit2:
Good for you then, if you reset everything in the UEFI to defaults you probably also reset some other overclocking settings that caused the instability. If the RAM frequencies are still set to the XMP profile and CPU set to stock frequencies, and you do not experience any BSODs in CE or Prime95 for a few hours, then you should be good to go. Else, if it crashes on auto frequency and auto voltage, RMA that chip.

[oronnoro]

I'll do the prime test just to see what it will do, if anything. If it happens again, I'll RMA it.

Again, thanks, take it easy and stay safe.