Return to

Help with a mystery Issue no one can explain


Dealing with a weird one.

X99 build:

Gigabyte X99 Ultra Gaming
32 Gigs Corsair Vengeance 3200mhz DDR4
Corsair RM750 PSU

Not overclocked
APC Surge protector

Built it in late 2016… ran fine for almost 2 years, then rebooted and wouldn’t post… RMA’ed mobo, Gigabyte tested fine

Tested PSU and voltages are fine also tested with another known working PSU

RMA’ed CPU and they replaced it, Intel never said if it was dead or not…

Put it back together and one of four sticks of ram is bad,… replaced that… worked for 8 months then bam same thing… bought used motherboard for testing CPU and same issue… GPU is fine as well… Co-worker tested all 4 sticks of ram on his new PC and they are good… I even have Fractal Design replacing the front panel of my case in case it shorted something out…

Trying to find the root cause not just throw money I don’t have at it to replace parts. Everything is under warranty except PSU. I just cant RMA again and get it back as good parts then find it happens again in 8 months… Everyone tells me CPUs rarely die yet I might have 2 on my hands with a still working board…

Round 2 of Gigabyte testing my mobo. It’s still good. I talked to the CS manager and he broke it down how detailed they get in testing and what tests they run, they seem pretty intense.

Trying to figure out what is killing my CPU w/o killing anything else




Mobo has latest firmware?



Sounds like PSU is faulty. If you get rouge currents it can fry CPU/RAM/MOBO. Try to change to another PSU for some time. See if anything comes up again. Had the same problem, the thing is though it’s really hard to tell since it can run fine for couple months straight and you would never know.



Take your motherboard outside the case,
lay it on a box, and run the system outside of the case.
You mentioned that the front of the case being replaced.
My suspicion is that the case could actually be the root cause of the issue.
Double check front panel usb ports on physical damage of the pins etc.
Damaged or shorted front usb ports can lead to all kinds of weird issues.

Or maybe a stand off or back io panel lip could also toutch the board causing shorts.

1 Like


My next suspicion would actually be the psu.
Because you have re-seated the cpu a couple of times already.

Also double check external io usb devices you have connected to the system.
Disconnect them all except for keyboard and mouse.

Next to that, when you managed to get it to boot.
Go into the bios and change the SA (system agent) Voltage to 1.2V.

Last but not least, eventually update the bios.
But make sure the system is running stable enough before doing this.
If you manage to get the system to work properly again,
a bios update shouldn’t be needed persee, but could solve some,
stability issues.
X99 was the first platform with DDR4 and it wasn’t the smoothest platform wenn it first came out.
However since your Gigabyte ultra Gaming board is a X99 GEN2 board,
most of the bugs with the bios should been iron’d out already.
Because Gigabyte’s first gen X99 boards bioses are a total disaster.



Its cursed.



The UPS is sine wave or stepped sine wave?

I guess you’re using a good power supply with APFC. If the APFC circuit is trying to adjust for a wave that’s not made for it might damage the PSU and the components connected to it. Also it’s not going to pop a fuse or a capacitor since the APFC circuit works within specs defined by the manufacturer but can “stress” the components past the PSU too much. That’s why you’re not seeing an immediate catastrophic damage but it adds up over time. This my theory about it, since you ruled out every other component in a in-depth scientific way, one at the time and you got reports from Gigabyte about the testing methodology.



Well …


1 Like


Did you apply a load to your PSU when you checked the voltages? Something like a lightbulb or whatever. Just measuring the voltage with a voltmeter tells you next to nothing, even then you might not find any faults, might have to get your hands on an oscilloscope to truly know if your PSU is sending dirty voltages.



Before getting an oscilloscope, it is cheaper to throw out the PSU and buy a new one (or two).



I used the PSU on a prior system, then built this system, ran for about 2 years then ran again for 8 months… tested voltages and they are all dead on…

Doubt its the PSU



Yes it does



How do you explain ONLY the CPU dying twice and nothing else is affected??



Yes Mine was from the refresh in mid/late 2016



I did all this multiple times… not a grounding issue… Case is fine and I just replaced the front as a precaution… not really a suspect. If it did short, the motherboard would of been affected and not just the CPU



Not sure how me as a end user can load a PSU and test… all I can check is voltages



No idea what the PSU is… its a corsair RM750… Again whats killing me is why nothing else is dying except the CPU… Motherboards are more sensitive to things than the CPU, at least thats what everyone keeps telling me “Oh CPU’s rarely die” … starting to not believe it…



Do you know for sure that the CPU is dying? Or is it simply that it won’t post/no video? I mean you did say intel never told you if the CPU was ok or not. Is there a reason you assume it’s toast? Putting it in another Mb or something? It looks like you have a spare used MB but have you ever got it to post with another CPU than the one you’re trying to troubleshoot?

Other than power problems (or physical damage to the pins/socket), which you seem pretty confident is not the case, I’ve never seen a bad CPU. Back in the day I had people bring MB/CPU’s to me a say the CPU was DOA only for me to review the CPU jumper settings and point out they messed up the voltages. /shrug. Once they started handling that automatically I’ve never handled a bad CPU personally.

Anyways, without a really clear idea of everything you’ve tried I can only suggest to eliminate possibilities one by one systematically. Including cables, and your surge protector. Make no assumptions.



I was referring to the surge protector you’re using. Is it an uninterruptible power supply with batteries in it or just a surge protector? Maybe you should try and avoid using it to eliminate one variable.



I second @MisteryAngel 's case suspicion above.

You say it’s fine, but are you 110% sure?

The fastest way to circular problem solving and stuff “no one can explain” is to make an assumption about one of the components that turns out to be untrue.

You RMA’d the board, CPU and changed PSU. The case is a common component across all of your failures. One of the only common components that remain. Taking gut feeling, emotion and assumptions out of it - it is a possible cause. I’ve seen it before.

That or your UPS.