[SOLVED] PC rebooting itself without warning, error or bsod. Possible memory or CPU defect.

Toggle sidebar Toggle sidebar
O

OGSnow

Reputable
Mar 24, 2016
148
1
4,695
Hi.

Second time my PC reboots itself without any warning whatsoever just freeze for a second and reboot. Prior to this i had 2 maybe 3 memory management bsod across 8-10 months. Then it boots back as if nothing happened.

While i'm sure other things can cause this, those 2 (cpu and ram) are the only culprits that given me reason to be concern.

My CPU is a i7 4770k and is quite old in around 5-7 years perhaps i can't even remember correctly.
i'm also dealing with an unknown issue (likely the CPU) that for no apparent reason all my games suffered a signicant FPS drop. Games were i'd get solid 55-60 FPS are now running on 30-40s lowering graphics helps VERY slightly but nowhere near satisfactory plus the game looks like crap on low.

And those BSOD i mentioned i believe the one behind is my ram. (16gb)

The second auto reboot that happened (minutes ago) and i notice one thing right before it auto rebooted, my windows theme would switch out to something that looks like a compatibility mode, my internet went down (there was a red cross on the internet icon on system tray) and i think (not sure cuz happened fast) something was up with windows action center (something failed i think) not sure like i said happened fast.

Haven't updated anything. All i changed recently was reset a cpu tweak i did on the bios for better performance.

I've got no other cpu or ram for testing. i don't have a cd-dvd reader. Well i do but it doesn't work so same as not having one. I don't even own a flash drive and no one i know have one. I live on a rural area. I was thinking on Memtest to see how my ram was doing but without being able to burn a cd or use a flash drive i can't do it.

I am planing a upgrade but it's still months away scheduled to be done at beginning of next year but if my pc gets worse i may be forced to rush it. that's how i always solve hardware problems, buy new ones throw old ones away cuz honestly trying to fix it is a pain in the ass and in the end you end up needing a new one regardless.

Sorry for the rant.

The self reboots are not often per se. The last one was about 2 weeks ago during a game. uplay had crashed and asked to close the game and log in again which i was going to do after a finishing a little thing, few seconds later, game froze, black screen, reboot. Ran the game again after boot and played for well over 3 hours without any issue.

This time happened while i was watching a video on kmplayer and playing something on an PSP emulator. What advice would any of you give to me? Is it worth trying to find the culprit when i'll end up changing every hardware at the end when i upgrade? Would upgrading sooner be best and forget this issue?

Thanks. And sorry if i sound arrogant.

Edit: My gpu is a gtx 1070 if needed. Running win7 64
 
Last edited by a moderator:
Solution
DSzymborski
Sort by date Sort by votes
Windows has a built in memory diagnostic tool. Hit the start button and start typing "windows memory diagnostic" until it comes up in the search results and then launch it.

It is a tricky thing to diagnose, could be something overheating so it would be worth trying a monitoring tool like open hardware monitor to see how hot the CPU & graphics cards are getting.
 
Upvote 0 Downvote
@Dugimodo

Did it. No problem was detected. Tho i'd wager this method is weak sauce compared to Memtest.

Overheating? I honestly doubt it. When my cpu bios tweak was on my cpu was reaching 81-94C under full load. I recently undid said tweak. Prior to the self reboot event. Now it's temp reach 65-73C max. My gpu never goes over 70C under load (with a 38-50s idle temp depending on weather). my HDD's when weather is really hot can reach 40C but usually is 28-31C. Got a 32-40C cpu idle temp. And mobo... i don't think i ever witness it go over 28C. EVER.
 
Upvote 0 Downvote


No power erros on windows. I downloaded hardware info. Sorry the ignorance but what you mean with "see if it drops when running"? See if it what drops? Power? Also, is "log all values for report" the logging you mean? Cuz something could happen mid-sensoring and dk if the app would still save the reading...
 
Upvote 0 Downvote
remove any overclock drivers, check power connections to the motherboard and GPU.
confirm your power supply is providing correct power levels.

generally, this will be a power condition problem. too much power being used by a device will trigger the motherboard power protection circuits. power supplies can also have protection circuilts.
overheating can cause the motherboard logic to reset the cpu.

in general, remove any bios overclock, update the bios and motherboard drivers, and remove any gpu and cpu overclock, monitor voltages in bios, check temps, clean out any dust from any fans in the system.
if you have a factory overclocked gpu, you might want to underclock it to see if it has an effect.

certain bugs can trigger a live kd session that will reboot your machine without warning. but there will be a .dmp file in the windows error reporting directory. you would have to search your c:\ drive for *.dmp files to see if this is the case.

you should provide the mini memory dumps so we can see why windows shutdown your system with a bugcheck.
 
Upvote 0 Downvote
@johnbl

Sorry mate but you might as well speak greek cuz that's pretty much how i understood 90% of what you said. As for the .dmp file i did find it. But found way too many. some dating back to 2013 (maybe residue of old installations or tagged along with the backup? idk) Most of it are uplaywebcore and steamassert something dmp the only dmp i found that was relevant was a memory dmp but was dated back to 2015. the most recent one i found was a svchost dmp which from 17 this month which is one day after the self reboot event.

Sorry for the ignorance but i just game. The bios tweak i did was through a google search and still was quite afraid of messing up something badly and in the end ended up undoing it. The mobo state is the same state that it was when installed. Nothing was changed (apart from the cpu turbo tweak that i ended up disabling it) or updated. If i recall correctly it was about 3 years ago more or less.
 
Upvote 0 Downvote
Hi.

Sorry guys i realize you can't help if i can't follow your instructions. My lack of knowledge is to blame but i did call the guy who does maintenance on my PC to check it out. And, as i suspected, everything is working as it should. It's been over a week since the last (second) self reboot and all's functioning as if it never happened. The only thing about voltage that i find a bit unusual is that at times, when i game for hours without exiting the game, HWMonitor says 108-110% power on the GPU but the thing is, that has been happening long before this issue appeared and google says it's normal for some cards to automatically overclock themselves if temperature is not too high or something along the lines.

I've decided. This issue is rather annoying. I think for the sake of my peace of mind i rather rush my upgrade and forget about it. Problem is, in my upgrade, i wasn't planning on changing my PSU. It is a AeroCool ATX 800W Strike X EN53754.

Is it best to change the PSU as well? It is only 1 year and half old. The PC stays on between 7-12 hours daily. Half of that time in rainy or stormy days.
 
Upvote 0 Downvote
when you game for a while, the gpu get extra warm, when it gets hot it will pull more power than the specifications allow. if the power is pulled from the motherboard pci/e slot the motherboard protection logic will kick in to protect the motherboard graphics slot from melting and catching fire. it does this by a reseting the cpu so you reboot.

turn up the fans on the GPU if you can or underclock the gpu to reduce power consumption.



 
Upvote 0 Downvote
@johnbl

I have no idea how i would do that. Based on my searches, this is a gpu built-in feature. It happens automatically i don't see anything to disable or control it in any way. Also, My temps on gpu and mobo are pretty fine. gpu never goes over 75C even during load of 100% usage at that point the fans are not even at max rpm and only about 60% according to HWMonitor and mobo, while i had said that had never witnessed it go over 28C ever, i was wrong. It was something else on the mobo tab of the app. The Mainboard do go as up as 45-48C under load. Neither 75C or 48C look like catching fire or melting temps to be honest.

About my PSU should i keep it and test it when i get the new hardware to see if anything happens or is it best to assume its faulty and change it altogether? Gave model and PC time usage in previous post.
 
Upvote 0 Downvote
provide one of the memory dumps, sometimes people have duplicates of overclocking software installed.
I can take a quick look at the memory dump to see why windows shut down the system.




 
Upvote 0 Downvote
This kind of random crashes often indicate problems with hardware. The most common culprit is the motherboard. But I've seen video cards going bad cause this too. But from what you are describing, I would strongly suspect the motherboard. It would be nice if you had another system that is good, that you can move/swap the video card with to rule out the video card as being the culprit. Without more hardware to swap and test with, it will be very hard to pinpoint the problem. I've had flake hardware like this before, but I usually have a couple of machines and boards to swap with to try to find the problem.
 
Upvote 0 Downvote


I hear ya. When i upgrade, nearly all main hardware will be changed. Only the PSU will stay. Tho i heard PSU is also capable of causing this so i don't know if i should upgrade and remain with the PSU to see if there'll be any problems with the new hardwares or add a new PSU to the list and assume my current's faulty.

Based on what i've noticed throughout a year or more, only the cpu and ram gave me reason to pause. The cpu in form of an unknown issue that inflicted a significant fps drop on my games. And the ram in form of 2 bsod in an year something about memory management.
 
Upvote 0 Downvote
memory dump shows a hacked/infected nt kernel
12 errors : nt (fffff80003a2d8a1-fffff80003a3299f)


and a hacked/infected user kernel
37 errors : win32k (fffff960000e67ff-fffff960000e6826)

the system bugchecked when the memory manager attempted to clean up a process and delete its virtual memory

running a hack tool:
~CheathappensTrainer2307
~CheathappensTrainer8272

You also have a thread running chrome that attempted to access a resource 77,000 times and could not. looking at the calls the stack is corrupted.

I would fix the corrupted windows core files.
then see if there was a chrome extension that had a problem or attempted to use one of the functions that was modified in the core windows files. if it did it tried 77k times and the stack blew up.

you have some other problems but you probably have all ready reinstalled by now.
(windows 7 )
on windows 10 you can run a dism.exe command to make most of the fixes, windows 7 will require a repair to fix the core windows files.

when you reinstall, be sure to update this driver:
Realtek High Definition Audio Function Driver
RTKVHD64.sys Tue Oct 22 05:31:34 2013
http://www.realtek.com.tw/downloads/downloadsView.aspx?Langid=1&PNid=14&PFid=24&Level=4&Conn=3&DownTypeID=3&GetDown=false

(it fixes a bug that messes up graphic drivers)

I would also update this driver
Realtek RTL8168D/8111D Family PCI-E Gigabit Ethernet NIC
\SystemRoot\system32\DRIVERS\Rt64win7.sys Thu Apr 21 23:58:00 2016
http://www.realtek.com.tw/downloads/downloadsView.aspx?Langid=1&PNid=13&PFid=5&Level=5&Conn=4&DownTypeID=3&GetDown=false

fixes network bugs that cause virus scanners and streaming software to mess up.

you also had a unexpected mix of usb 3 drivers. one was 3 years older than the bios or the other two. see if you can update directly from intel:
Intel(R) USB 3.0 Host Controller Switch Driver
https://downloadcenter.intel.com/

chrome was using an extreme amount of pagedpool (c:\pagefile.sys)ie virtual memory

I think the system was waiting for a response from this driver
\SystemRoot\system32\DRIVERS\iaStorF.sys Wed Nov 4 02:27:52 2015
the system returned a error but the error was not processed correctly and the system got blocked up waiting for the request to be processed.

see if you can update iastorf.sys and iastora.sys from intel:
https://downloadcenter.intel.com/
 
Upvote 0 Downvote
@johnbl

I'm afraid i don't know how to proceed.

Firstly, i haven't reinstalled windows.
And about chrome, idk. I started using chrome only for a few months now. Got tired of FF's slow and high cpu usage.
The only extensions i have on chrome are: Video DownloadHelper, Google Docs Offline, Free Download Manager, Session Manager and Adblock Plus (fairly recently installed). On the CheathappensTrainer, I trust the tool. I doubt it's harming my PC directly. Maybe a conflict with some files or is that what're you saying?

Honestly, right now, i'm not looking for a fix for this issue cuz i've already made preparations and will rush the upgrade to happen sooner. My main concern right now is, Like i said before, the PSU is the only main hardware i do not intend to change. But my searches said it is capable of causing issues of this kind. So i wanted to know if i should stay with my current PSU, upgrade, and see how things will be, or assume it's faulty and get a new one. As i said prior, the model is a AeroCool ATX 800W Strike X EN53754 and about a year and half old.

Also, since i'm no longer looking for a fix for this system, would still be nice if you or anyone could list what could cause this in case others having the same issue stumble upon this thread.
 
Upvote 0 Downvote
I think the root cause of this memory dump problem will be the old intel storage driver not correctly processing a cancled request. the cancel was pending but the driver thought it had nothing to work on.
----------
you should also download
https://docs.microsoft.com/en-us/sysinternals/downloads/vmmap
and run it to see the problem. you can find the program that is using the excessive pagefile space then select the empty working set menu.
it will either reset the working set or your system will bugcheck


----------------

psu would not be the problem. power supplies do not target key windows data structures and make planned changed. psu problems are more random.

do try to update the intel storage driver or remove it and go to the Microsoft storage drivers. The Microsoft storage driver is updated via windows update. you have to manually update the intel storage driver. because of the error in the storage driver adding to virtual memory was working but deleting was failing so the app keep using more and more pagefile.sys space.

--------

see the note about updating iastoref.sys and iastorea.sys from intel.
i think it is why chrome got messed up.



 
Upvote 0 Downvote
You don't really *need* to replace the PSU. This one is from one of Andyson's non-terrible platforms, so you lucked out in that it's better than the typical Aerocool PSU (it's not a brand synonymous with quality). You did kind of prioritize wattage out you weren't even close to needing (this system would run fine on a quality 550W) over quality, however, and I'd recommend you go a different route next time you need to replace a PSU.

Even if you don't think the trainer is the issue, for testing purposes, you should be suspicious of *anything* you add beyond the barebones Windows install for a problem like this. You'll find yourself misdiagnosing certain issues if you just assume something is OK. And I'd still try and reinstall Windows, there's no reason to let what may still be a perfectly good spare PC rot when there are simple things to do; just wipe the machine and do a fresh install.
 
Upvote 1 Downvote
@DSzymborski

Yea i didn't really thought it through when i bought the PSU back then all i knew was that i needed a more powerful PSU for my gtx 1070 (i had a 400-450W i think). So yea, next time i will put the same amount of effort i did on other hardwares and prioritize quality over more watts. Also on that, i'm out of the loop in terms of psu brands. Could you recommend me one for the future?

On the trainer, oh yeah, trust me, i do suspect of things and too much most of the time (natural paranoid person here). But i've been using the tool for over 5 years and never had any issues whatsoever. No way i'd leave that on my PC if i didn't trust it.

I value my piece of mind greatly. So, on my upgrade i was planning on getting 32 of ram which may be a bit of overkill even for a heavy gaming pc but rams are so expensive here on my country that i can very well get 16 and save enough money for a quality psu juuust in case my current one ends up in need of replacing in the future (hopefully don't but having a plan b is always reassuring).

I will not let my system rot per se. all the spare hardware won't be thrown away. Will likely give parts that won't use for making a rig for family member. So, some windows reinstalling will happen but will no longer be my issue. I was already planning on this upgrade ever since my cpu decided that, for no logical reason, strike my games with 8-10 FPS drop that happened over a day that the only thing that changed was that a game was updated through steam which makes 0 sense to me. (did try completely uninstalling said game but fruitless). And even if that wasn't the case. i'd eventually, maybe in a 6 months time i'd upgrade anyway. modern AAA titles are demanding. This issue merely making the upgrade to happen much earlier.

Thank you for the response.
 
Last edited by a moderator:
Upvote 0 Downvote


Generally speaking, for a PSU on a budget, Corsair's newer CX series with the gray-and-black label (not the older ones with the green letters) are competent budget choices. For a higher-tier of a PSU, generally EVGA G2/G3/GS/P2/T2 or SeaSonic Focus/Prime series or Corsair RMx are top-tier, easy to find, and with brands having a *lot* of models, there's usually a good sale on one (my EVGA G3 was only $69.99 on sale, which was an absolute steal).

There are other good PSUs out there, of course, but on a very general recommendation it's hard to go wrong with one of these.
 
Last edited by a moderator:
Upvote 1 Downvote
Solution
@DSzymborski

Checked the place where i buy my hardwares. Tons of model from those brands available from lower to higher end it seems.
Thank you. Next time i buy a new PSU will make a smart purchase and not just cuz it has more watts.
Hopefully that's considered a hookie mistake haha.
Cheers man!
 
Upvote 0 Downvote
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom