[H] community I am begging for your help.

zamardii12

2[H]4U
Joined
Jun 6, 2014
Messages
3,409
So, my computer which is in my sig has been running fine since I put it together however starting this past Saturday has been absolute frustration. I am a techy person and I cannot for the life of me figure out what is going on with this thing. It started with random freezes... now, since the computer is mostly used for gaming and regular typing up school stuff it doesn't see much variety in tasks. In any case, I was playing a game and it freezes or reboots... here is a small list of what i would do after any random amount of usage...

  • Audio stutter and display freeze forcing me to pull the electric plug (or hold down power button)
  • Screen goes black in the middle of a game, display goes to power save mode, and then reboot after a few seconds
  • Screen goes black, monitor goes to power save mode, and computer is stuck doing whatever
Now to clear a few things up:

  1. Nothing is overclocked
  2. CPU was overclocked slightly to 4GHz (very mild) but was reset to default once issues started
  3. No mining ever done on this computer
  4. Computer was always plugged into a surge protector and battery backup
I brought the computer into work after spending 2 days futzing with it with no results. Completely re-downloaded Windows, didn't even update to the latest Nvidia drivers or anything... I am just on whatever Microsoft found through updates, and literally just now the computer again rebooted on its own. There are no beeps, or error codes on the motherboard that could point me in the right direction; At various points I thought it was the videocard, or SSD the OS is on, thinking maybe the system isn't getting enough power delivery but it's a 850w Gold PSU from EVGA... but there is no explanation for this behavior.

Does anyone have any idea what else I could do at this point? I was thinking of just buying a new motherboard and seeing what happens, but I am just not happy with that solution. Also thinking maybe replacing the power supply could do it? Just not sure. Any advice at all would be greatly appreciated.

Btw, I am getting Kernel Power issues. See below.


jHQ3Uya.jpg


kiYXB0o.jpg
 
Last edited:
Did you run memtest? RAM issues can cause weird problems.
You have another GPU to test with?
Maybe it's Power delivery problem? First thought on that is PSU like others mention, but it could also be problem with your case! I had a similar problem some time ago.
My issue was either reboot or shutdown at random times (if I remember symptoms correctly). I tried replacing various parts with stuff I had on hand, but in the end it turned out to be a malfunctioning power button that I had to rewire. It was very strange. You might check and see what your settings are for when you press your power button. I know chances are slim, but hey... maybe a good place to start.

What's left after that?
Motherboard?

I'm not much help am I?
Anyway good luck. Curious to know how it all works out.
 
Check the PCIe power cables to your graphics card. I had some get a bit burned and cause a bad connection, leading to exactly the symptoms you describe.
 
This is what I would do if this happened to me.

BTW I also have the KS8000 49" and it's incredible. I also used to own the 5820K and loved that as well.

1 ) Backup all the files you want to keep. Then move to step 2.

2 ) Pull your panel off and re-seat your CPU. Even if this means taking off the water pump / Air Cooler which is very easy to do. Re-seat your memory. When you do this, move the sticks around. All of your power connectors from, from the PSU and motherboard, re-seat them. This takes a few minutes. Basically go through your system and re-seat everything. Don't rush this job, you want to make sure it's done well as to eliminate this being one of the issues.

Memory normally just doesn't go back. Especially if you have this under a sine-wave battery backup. You're getting good clean electricity. 4ghz is also a mild OC. No concerns there.

3 ) Re-install Windows 10. Try and use your motherboard disc to update all the drivers. If you can, get back to the basics. Windows 10 will attempt and be able to update some drivers. Let this happen I guess.

Throw on a few games and see what happens.

I personally think it's software / bad connection.
 
Others have already given some good advice, so I will list an obscure one just in case you tried all of their advice and nothing's changed.

Check that your reset or power header cable from case to motherboard isn't pinched between something and your case, and for that matter, check if any loose cables are pinched.
 
i had a weird issue like this but i also had some other blue screen. I had a core going bad.
 
Did you run memtest? RAM issues can cause weird problems.
You have another GPU to test with?
Maybe it's Power delivery problem? First thought on that is PSU like others mention, but it could also be problem with your case! I had a similar problem some time ago.
My issue was either reboot or shutdown at random times (if I remember symptoms correctly). I tried replacing various parts with stuff I had on hand, but in the end it turned out to be a malfunctioning power button that I had to rewire. It was very strange. You might check and see what your settings are for when you press your power button. I know chances are slim, but hey... maybe a good place to start.

What's left after that?
Motherboard?

I'm not much help am I?
Anyway good luck. Curious to know how it all works out.

This is what I would do if this happened to me.

BTW I also have the KS8000 49" and it's incredible. I also used to own the 5820K and loved that as well.

1 ) Backup all the files you want to keep. Then move to step 2.

2 ) Pull your panel off and re-seat your CPU. Even if this means taking off the water pump / Air Cooler which is very easy to do. Re-seat your memory. When you do this, move the sticks around. All of your power connectors from, from the PSU and motherboard, re-seat them. This takes a few minutes. Basically go through your system and re-seat everything. Don't rush this job, you want to make sure it's done well as to eliminate this being one of the issues.

Memory normally just doesn't go back. Especially if you have this under a sine-wave battery backup. You're getting good clean electricity. 4ghz is also a mild OC. No concerns there.

3 ) Re-install Windows 10. Try and use your motherboard disc to update all the drivers. If you can, get back to the basics. Windows 10 will attempt and be able to update some drivers. Let this happen I guess.

Throw on a few games and see what happens.

I personally think it's software / bad connection.

Others have already given some good advice, so I will list an obscure one just in case you tried all of their advice and nothing's changed.

Check that your reset or power header cable from case to motherboard isn't pinched between something and your case, and for that matter, check if any loose cables are pinched.

i had a weird issue like this but i also had some other blue screen. I had a core going bad.

So I haven't had time to re-seat everything on the system however I did call EVGA (the makers of the PSU) and described everything to the person I talked to and he said it definitely sounds like a PSU issue. So, they setup a RMA for me and i'm going to send the PSU back. If I get the PSU back and I have the same issues I will re-seat everything on the system, remove and reattach every cable going to the motherboard. If the problems persist then I am going to go with a motherboard issue, however I am fairly hopeful the PSU is the culprit here and getting it replaced/fixed will fix my issues.

The turn-around for the PSU at this rate doing a standard RMA will be around 2 weeks so I will update once that's complete and I have had some time to test everything once I get the replacement.
 
Check the PCIe power cables to your graphics card. I had some get a bit burned and cause a bad connection, leading to exactly the symptoms you describe.
The good old molex PCIE power adaptors.
 
Check the PCIe power cables to your graphics card. I had some get a bit burned and cause a bad connection, leading to exactly the symptoms you describe.

Really? A bad connection to a videocard power-cycled your computer? That doesn't make sense. Not doubting you I just don't see why that'd happen. I would just suspect a black screen and the overall power to still be on.
 
Really? A bad connection to a videocard power-cycled your computer? That doesn't make sense. Not doubting you I just don't see why that'd happen. I would just suspect a black screen and the overall power to still be on.

Sorry, to clarify, it didn't actually power cycle, but froze and I had to hard reset it, so effectively the same thing.

I suspect in your case though that it's simply a bad PSU. It happens.
 
This is more than likely a power supply problem. However, it could be a number of other things. I've seen this happen on my own systems over the years on rare occasion. It was always related to power in my case. In most of the cases it was because the 900watt output of my UPS was insufficient. They would alarm and show that the system draw was around 950watts or so. This was always when I had three graphics cards in some more power hungry configurations than I run today. Additionally, I've seen this very problem happen due to video card issues. Memory issues are also potentially common in these cases. I've seen that on other people's systems.

As for the commentary about reseating everything, I'm going to tell you not to waste your time. Your CPU is seated or it isn't. The socket design doesn't allow for "partial" connections whether we are talking about LGA or ZIF sockets. If a CPU wasn't making good contact, it generally won't work at all or you would be missing something like memory channels. RAM is easy to see if its making contact. Either you've got all the memory installed reporting in BIOS and Windows or you don't. Reseating cables often makes sense as these connections have weight of the cable pulling on the connection or the connection is flaky because they heat up, cool down, or aren't attached properly. If a graphics card weren't seated properly it would either not work or show x8 PCIe lanes instead of x16 or x4 instead of x8.
 
Save yourself some trouble by looking at the capacitors on your motherboard. If any are swollen or leaking (even a tiny bit) the board should be replaced.
Of course, you are getting an RMA power supply so go with that too.
 
Save yourself some trouble by looking at the capacitors on your motherboard. If any are swollen or leaking (even a tiny bit) the board should be replaced.
Of course, you are getting an RMA power supply so go with that too.

True. That is a good point.
 
Which EVGA PSU specifically, they had some issues with the G2 series I believe. Try running with minimal power connection as well, could be a bad cable or device shorting.
 
Last edited:
I'd start off by switching out the PSU and seeing what happens....

After that, the motherboard.

It bound to be one of those two components.

Sounds like a power supply issue to me.

Did you run memtest? RAM issues can cause weird problems.
You have another GPU to test with?
Maybe it's Power delivery problem? First thought on that is PSU like others mention, but it could also be problem with your case! I had a similar problem some time ago.
My issue was either reboot or shutdown at random times (if I remember symptoms correctly). I tried replacing various parts with stuff I had on hand, but in the end it turned out to be a malfunctioning power button that I had to rewire. It was very strange. You might check and see what your settings are for when you press your power button. I know chances are slim, but hey... maybe a good place to start.

What's left after that?
Motherboard?

I'm not much help am I?
Anyway good luck. Curious to know how it all works out.

test/change your cmos battery, try another psu then a gpu.

Check the PCIe power cables to your graphics card. I had some get a bit burned and cause a bad connection, leading to exactly the symptoms you describe.

PSU then mobo

PSU problem.

The good old molex PCIE power adaptors.

This is more than likely a power supply problem. However, it could be a number of other things. I've seen this happen on my own systems over the years on rare occasion. It was always related to power in my case. In most of the cases it was because the 900watt output of my UPS was insufficient. They would alarm and show that the system draw was around 950watts or so. This was always when I had three graphics cards in some more power hungry configurations than I run today. Additionally, I've seen this very problem happen due to video card issues. Memory issues are also potentially common in these cases. I've seen that on other people's systems.

As for the commentary about reseating everything, I'm going to tell you not to waste your time. Your CPU is seated or it isn't. The socket design doesn't allow for "partial" connections whether we are talking about LGA or ZIF sockets. If a CPU wasn't making good contact, it generally won't work at all or you would be missing something like memory channels. RAM is easy to see if its making contact. Either you've got all the memory installed reporting in BIOS and Windows or you don't. Reseating cables often makes sense as these connections have weight of the cable pulling on the connection or the connection is flaky because they heat up, cool down, or aren't attached properly. If a graphics card weren't seated properly it would either not work or show x8 PCIe lanes instead of x16 or x4 instead of x8.

Well everyone I am back. Got the new PSU from EVGA (replacement) in the mail today. Hooked everything back up, powered on, was on for maybe 3 minutes and then white screen and reboot. Event Viewer shows the same Kernel-Power Error Code 41 as before. I have just about had it. Was really hoping this would fix this issue. What's next? Replacing the motherboard? I downloaded MEMTEST but I somehow doubt that's the problem.
 
The error message is so generic this is one of the more difficult things to troubleshoot. Is it possible to try with onboard video and pull out the graphics card? You may have to eliminate components, such as USB devices to video cards to maybe even memory sticks. Trial by elimination of anything that maybe connected. I'd start with graphics and use onboard and see if there is any luck.

Blah
 
Well everyone I am back. Got the new PSU from EVGA (replacement) in the mail today. Hooked everything back up, powered on, was on for maybe 3 minutes and then white screen and reboot. Event Viewer shows the same Kernel-Power Error Code 41 as before. I have just about had it. Was really hoping this would fix this issue. What's next? Replacing the motherboard? I downloaded MEMTEST but I somehow doubt that's the problem.

I bet you need a different power supply, or you have a bum cable/connector. Or you could have a ground.

Take your board out and put it on a cardboard box. Install a linux live disk on a thumb drive. Hook up 1 memory stick, and your video card, keyboard/mouse. Nothing else. Boot up to the memory stick and see how it runs. If it's not crashing, do a stress test.

Honestly, I would get a different psu if it does.

Edit: Also check your v-bridge for anything that looks out of place. Swollen caps/etc on the board.
 
Last edited:
giphy.gif


Pretty much everything people said above ^ .

Time to play the elimination game.

First off yeah, go ahead and run Memtest, then yank out the graphics card and other exp cards individually, and then if none of that narrows it down....

I don't know how many sticks of RAM you have, but alternate how many and to different slots etc etc (always fun to play RAM musical chairs smh)

If none of that does anything, only then should you bother going case-less to see if your case is somehow shorting something.

Andddddd, if none of that presents a solution: it's the board..
 
I bet you need a different power supply, or you have a bum cable/connector. Or you could have a ground.

Take your board out and put it on a cardboard box. Install a linux live disk on a thumb drive. Hook up 1 memory stick, and your video card, keyboard/mouse. Nothing else. Boot up to the memory stick and see how it runs. If it's not crashing, do a stress test.

Honestly, I would get a different psu if it does.
.

I get that, but it ran solid for 2 years with this PSU and I don't have that many things in my system. I just got a brand new 850 Watt Gold PSU. If the PSU is the problem that'd be very strange since it ran perfectly fine for the longest time.
 
Which EVGA PSU specifically, they had some issues with the G2 series I believe. Try running with minimal power connection as well, could be a bad cable or device shorting.

It's the EVGA SuperNOVA 850 G2, 80+ GOLD 850W 220-G2-0850-XR
 
OP said he keeps his rig plugged into a battery backup.

So the G2 brownout stuff is a non-issue.

-- At this point, I'd bet good money its the motherboard.. Just saying
 
View attachment 89493


First off yeah, go ahead and run Memtest, then yank out the graphics card and other exp cards individually, and then if none of that narrows it down....

Just got done running MEMTEST for 2 and a half hours and not one single error. I ran every test 4 times each and not a single error. I am going to try switching around memory tomorrow.
 
Could it have something to do with the power settings in Windows? Make sure it's set to max performance. You could also try a CMOS reset.
 
Oh boy, an online friend has a similar issue, and its a PITA the debug.

If possible, save your crash mini-dumps (C:\Windows\Minidumps) and post them. I (or someone else, or you can PM me) can use windbg to get a stack trace to see if there's any software component in common. If not, then its a hardware issue; but knowing what precipitated it will narrow it down.

Just because it says 'Kernel-Power' doesn't mean its a PSU issue. I mean, FFS, it could be totally different bug checks.

Edit: Assume minidump will reveal all your apps running at time of crash (its whatever process that called the bug check), FYI.
 
Last edited:
I think the horrible truth is that it's simply the motherboard.

One more thing to try though, try running your Graphics card PCI x16 Express port at Gen2, rather than Gen3, and see if that makes a difference. It changes more than just the bandwidth, as it also affects signalling etc, and I necessarily have to do this on my rig where I've overclocked my BCLK alot.

I realise you're not overclocking, but worth a try to see if it makes a difference...


I should add that I originally didn't have to do this. I don't know if it's the difference between going from a Titan XP to a Titan Xp CE, but I now necessarily have to use PCI 2.0 to be stable.
 
I think the horrible truth is that it's simply the motherboard.

That's what i'm starting to think in which case i'll just get a i7-8700 and abandon X99 and my 5820k.

The strange thing is I brought the computer into work ran Prime95 on it for an hour or more and now I am running 3DMark on a loop and not one crash or reboot yet. Still running vanilla windows with nothing extra installed except Steam and OpenHardwareMonitor.

It's a shame that this problem is SUCH a pain in the ass to pinpoint, but I do hear a clicking-like noise every once in a while and it's not coming from the PSU. I think the motherboard is just done and I don't know how or where. At least everything I have will work on the new motherboard and CPU including the memory so I just need to replace those two components.
 
That's what i'm starting to think in which case i'll just get a i7-8700 and abandon X99 and my 5820k.

The strange thing is I brought the computer into work ran Prime95 on it for an hour or more and now I am running 3DMark on a loop and not one crash or reboot yet. Still running vanilla windows with nothing extra installed except Steam and OpenHardwareMonitor.

It's a shame that this problem is SUCH a pain in the ass to pinpoint, but I do hear a clicking-like noise every once in a while and it's not coming from the PSU. I think the motherboard is just done and I don't know how or where. At least everything I have will work on the new motherboard and CPU including the memory so I just need to replace those two components.

My immediate reaction to that is try a different room and power outlet at home...
 
You could also try running a Live Linux install for a while, to test for software issues.
 
  • Like
Reactions: spine
like this
Try running just 8GB in dual channel.

If that works it would indicate the CPU/mobo is at fault.
 
Had a similar problem that eventually turned out to be a failing 4TB spinner drive that was being used for storasge. Computer finally stopped booting until after I removed the drive!
 
Just got done running MEMTEST for 2 and a half hours and not one single error. I ran every test 4 times each and not a single error. I am going to try switching around memory tomorrow.
Wait, hold on. You mean the system didn't reboot or shutdown or have any issues for 2 hours? Did you run memtest inside of windows or was it a stand-alone boot of some sort? What hardware did you have connected during the test--gpu? hard drives?

If it's solid for 2hrs of memtest, I don't think there's anything fundamental wrong with the hardware in terms of the cpu motherboard or memory.
 
Wait, hold on. You mean the system didn't reboot or shutdown or have any issues for 2 hours? Did you run memtest inside of windows or was it a stand-alone boot of some sort? What hardware did you have connected during the test--gpu? hard drives?

If it's solid for 2hrs of memtest, I don't think there's anything fundamental wrong with the hardware in terms of the cpu motherboard or memory.

Yeah I booted up into MEMTEST so I tested outside of Windows. It ran the whole test 4 times. Afterwards when I booted into Windows I ran prime95 for 4 hours and then Firestrike in 3DMark on a loop for 4 hours. Afterwards I kept the computer on for 3 days straight and it never turned off. I am completely perplexed what the hell is going on. So we'll see what happens this weekend if anything.
 
Back
Top