Question Brand new system build, horribly unstable, extensive testing, no where left to turn.

[level42]

*Cautionary note, my stress levels are dangerously high right now, I feel like I'm on the verge of a heart attack over this entire adventure, please excuse any incessant ramblings ...

I recently decided to build myself a new computer, after close to 10 years.

Parts list below:

NEW:
Motherboard: GIGABYTE Auros x570s Master (Rev 1.0)
CPU: AMD 5950x
**RAM: G.Skill 32768 MB (DDR4-2666) - XMP 2.0 - P/N: F4-3600C18-32GTZN - QTY: 4
**SSD: Samsung 980 Pro (2TB NVMe)
Cooler: Kraken x73

** Likely not at fault, see Troubleshooting Steps below.

RECYCLED FROM OLD PC:
OS: Windows 10 Enterprise x64
PSU: Seasonic X series GOLD (1250w)
GPU: MSI 1080Ti
Chassis: Fractal Define R6

Proper parts link here: https://valid.x86.fr/697xnt

Issues:
Almost immediately upon first boot, the computer exudes bizarre behavior. Consistent, steady, chaotic random crashes. Primarily BSOD's within Windows.

Numerous different stop codes, some examples below:

• KMODE_EXCEPTION_NOT_HANLDED
• DRIVER_IRQL_NOT_LESS_OR_EQUAL (ACPI.sys)
• CACHE_MANAGER (Ntfs.sys)
• IRQL_NOT_LESS_OR_EQUAL
• SYSTEM_SERVICE_EXCEPTION (NETIO.SYS)

* Among several other variations of the same messages, all seeming reporting something different, no consistency in any of these BSOD's.
** Numerous other crashes as well, of which I don't have records since I've reformatted numerous times now.
** * We're talking within an hour, I might get close to 20 crashes ...
**** Sometimes, when it successfully boots into Windows, I can use the system for maybe a few hours, then all of a sudden a single BSOD, which seems to then trigger a PLETHORA of BSOD to promptly follow. Almost as if the crashes are daisy chained together. (Not sure if this is valuable info or not)

Troubleshooting steps:

• Reinstall Windows 10 (Repeated several times)
• Remove SSD from equation and switch back to older SATA SSD (To me, proves the SSD is not at fault)
• Ran MemTest for 48h on all 4x sticks (Found 2 errors, report didn't save, but I have photos of the issues here).
• Installed a single RAM module in slot A2 (Per Manual) - Rotated out each memory module 1 by 1
• Re-seated each and every component (RAM, CPU, GPU, SSD, Motherboard, etc)
• Flashed, and re-flashed BIOS numerous times between F2, F3, and F4b
• Enabled/Disabled X.M.P on memory
• Installed ALL windows updates, and ALL drivers per the manufactures (GIGABYTE Desktop app helps here)
• Confirmed part compatibility vs. QV's
• Probably others which I can't recall over my 2 - 3 weeks troubleshooting this.

Seemingly, no matter what I attempt, nothing, at all helps. I've got several thousand dollars invested into this build, and I effectively have absolutely nothing to show for it. I've already reached out to GIGABYTE for some initial support on this Motherboard, but let me just say how their support channels, sucks a bag.

Small rant digression (Can be skipped):
My only option with GIGABYTE, is to submit a ticket. Which takes 5 - 7 Business days before they'll even respond to me, with what I can only assume is some cut and pasted basic troubleshooting steps. After which, when I reply, have to then wait yet another 5 - 7 business days for a response. I cannot find a phone number to call anyone, nor an email to reach anyone. Further to this, once the ticket has been submitted, you cannot modify the ticket, or append any new information to the ticket, since you're troubleshooting your issues on your own, and discovering new information while you wait 5 - 7 business days. To me, this is just unacceptable, and this alone has steered me to never want to buy another GIGABYTE product in the future.

/End rant​

​

At this point, I'm at a complete loss here. I have all of these parts, and can literally not do anything productive with them. I've exceeded the 15 day in store return policies on all parts at this point, and I'm effectively stuck with a pile of very, very expensive paper weights.

Where I need help:

At this point, I am completely out of ideas one what to do/test. I'm looking for advice on further things that I can test in house, to maybe determine the root cause of this nonsense. It's become a literal nightmare for me, and even if I manage to "fix" this, I'm terribly uneasy about the longevity of this entire build. I bought these parts with the intent on them being "future proof" and lasting me another 10 years before having to buy another computer.

I do not own any other "cross compatible" parts, so I can't put the CPU in another board to test, nor vise-versa either. I'm at the mercy of testing the equipment, as is.

I feel a bit out of my league here, as technology has evolved and changes dramatically, since I last built myself a system (Intel i7 3770k era)

If anyone can provide any assistance or guidance, or advice in any way shape or form, please let me know. I will forever be in your debt.

Thank you!

 

[punkncat]

Based on the commonality of system memory in most of those stop codes and the fact that memtest returned with an error would seem to be indicative of an issue with the memory.
I cannot imagine a situation where all 4 of 4 sticks would be bad, which you seem to indicate that you tested each stick individually and still experienced the error.

This leads me back to CPU in this case as that is where the memory controller is. It wouldn't for sure rule out motherboard in my eyes either. I would check the CPU to be sure you didn't bend or break a pin. Aside from that it is pretty much a guessing game as to which of the three are likely the issue.

I would make the assumption that you are monitoring temperatures and have correct cooler install. Don't think that's the issue but worth mentioning.

This is a tough one. Were these items purchased online, or at a brick and mortar? Are you within return or RMA period?

 

[kerberos_20]

memory controller and 4 dual sided sticks running at 3600MHz....well its not impossible to run that, but you will need to raise voltage for memory controller/infinity fabrics or reduce ram clock, ryzen 5000 serie has official clock at 3200MHz, going above that will need some voltage tweaks as 3 things gets overclocked at once - memory controller, infinity fabrics and ram itself


you can do stress test to figure out your stable settings faster, test will fail much quicker (within seconds/minutes) compared to memtest
here use aida64 ram stress test

 

[level42]

Based on the commonality of system memory in most of those stop codes and the fact that memtest returned with an error would seem to be indicative of an issue with the memory.
I cannot imagine a situation where all 4 of 4 sticks would be bad, which you seem to indicate that you tested each stick individually and still experienced the error.

This leads me back to CPU in this case as that is where the memory controller is. It wouldn't for sure rule out motherboard in my eyes either. I would check the CPU to be sure you didn't bend or break a pin. Aside from that it is pretty much a guessing game as to which of the three are likely the issue.

I would make the assumption that you are monitoring temperatures and have correct cooler install. Don't think that's the issue but worth mentioning.

This is a tough one. Were these items purchased online, or at a brick and mortar? Are you within return or RMA period?

I did a through inspection of the CPU pins when I reseated the CPU and they all looked flawless to me. Everything was perfectly straight.

All parts except the cooler are from Brick and Mortar, but I'm past the 15 days return/exchange period unfortunately. I'm well within my RMA period, but I cringe at the idea of sending an item back "just because" and having them simply mail it back to me saying they couldn't find fault.

memory controller and 4 dual sided sticks running at 3600MHz....well its not impossible to run that, but you will need to raise voltage for memory controller/infinity fabrics or reduce ram clock, ryzen 5000 serie has official clock at 3200MHz, going above that will need some voltage tweaks as 3 things gets overclocked at once - memory controller, infinity fabrics and ram itself


you can do stress test to figure out your stable settings faster, test will fail much quicker (within seconds/minutes) compared to memtest
here use aida64 ram stress test

I appreciate the suggestion, I will attempt it tomorrow evening after work and see how far I can get. Thanks!

I guess the X.M.P Profile on the chip is for the "sold" amount? Which in this case would have been for running them in a pair, since they were bought in a pair? Would this be a fair assumption? Granted, I still experienced crashes when running the sticks as a pair only, with two empty slots in both X.M.P enabled and disabled.

 

[helper800]

memory controller and 4 dual sided sticks running at 3600MHz....well its not impossible to run that, but you will need to raise voltage for memory controller/infinity fabrics or reduce ram clock, ryzen 5000 serie has official clock at 3200MHz, going above that will need some voltage tweaks as 3 things gets overclocked at once - memory controller, infinity fabrics and ram itself


you can do stress test to figure out your stable settings faster, test will fail much quicker (within seconds/minutes) compared to memtest
here use aida64 ram stress test

My 3900x is running 4x8 gb 3600mghz cl14 kit of G skill RAM. The vast majority of AMDs CPUs since 3000s can handle those timings and frequencies, especially if it is an XMP profile. It could also be the motherboard itself though. There does seem to be some sort of memory subsystem issue. In order of likelihood in my opinion; RAM, Mobo, CPU. Start with running memtest with non xmp base speeds all 4 sticks. If you get a single error I would get a new RAM kit but keep old sticks. If this solves your issue them RMA the old kit. We can go from there if this doesn't solve the issue.

 

[kerberos_20]

My 3900x is running 4x8 gb 3600mghz cl14 kit of G skill RAM. The vast majority of AMDs CPUs since 3000s can handle those timings and frequencies, especially if it is an XMP profile. It could also be the motherboard itself though. There does seem to be some sort of memory subsystem issue. In order of likelihood in my opinion; RAM, Mobo, CPU. Start with running memtest with non xmp base speeds all 4 sticks. If you get a single error I would get a new RAM kit but keep old sticks. If this solves your issue them RMA the old kit. We can go from there if this doesn't solve the issue.

thats a single rank ram, about same as two dualsided modules

 

[helper800]

thats a single rank ram, about same as two dualsided modules

The amount of ICs / ranks of the ram typically does not significantly contribute to issues or instability with RAM when there are "too many" (4 x DualR modules). Nowadays a lot of 16 gb DDR4 modules are single rank anyways. Unless you know something I do not, which I will admit is very possible, I do not understand how this could be the issue.

 

[level42]

So, I'm no major expert here, but from what I can tell, and based on at least bits and pieces of the advice I've received thus far from this community. The likely suspect is something, in and around the memory. Now, I find the likely hood that I received 4x memory modules, all dead, to be very unlikely. This is not to say that at least some of the memory modules aren't bad, but I just don't think all 4 are D.O.A. Thus, I'm hung up on the idea that the motherboard is DOA.

I've got a support thread open with GIGABYTE at the moment, and they've since replied back with some "additional troubleshooting" steps, such as check CPU for damages, update drivers, use alternate boot device, etc. All things I've already examined and tested whilst waiting for their initial reply (Great support GIGABYTE, get it together)

My hope is that if I get the motherboard RMA'd, then at the VERY least, I can likely rule it out as problematic (if the problem persists, which for the sanctity of my own heath, I hope is NOT the case.), but frankly, I feel like there's really little to nothing else I can do at this point, short of just throwing even more money that I don't have at this thing.

I'm still open to additional suggestions if anyone has anything else they thing might be the case.

Per the issues/idea around the RAM needing additional voltages to run properly, while this "may" be the case whilst using 4x modules, I don't think this is the root cause, as I have the exact same issue running any of the 4x modules as a single module in the PC (Running with only 1 stick, vs 2 or 4). Since the theory was around running the 4x sticks, I'm not fully convinced that I need to test this any further at the moment, until I can confirm stability with a single stick first. (Correct me if I'm wrong, and please remember that OC'ing components is not my forte , at all.)

 

[helper800]

I'm still open to additional suggestions if anyone has anything else they thing might be the case.

Per the issues/idea around the RAM needing additional voltages to run properly, while this "may" be the case whilst using 4x modules, I don't think this is the root cause, as I have the exact same issue running any of the 4x modules as a single module in the PC (Running with only 1 stick, vs 2 or 4). Since the theory was around running the 4x sticks, I'm not fully convinced that I need to test this any further at the moment, until I can confirm stability with a single stick first. (Correct me if I'm wrong, and please remember that OC'ing components is not my forte , at all.)

Try a random stick in slot 4 from the CPU socket. Go into BIOS and make sure that XMP is off and set voltage to 1.35. Save and exit. Run memtest.

 

[level42]

So, quick update.

I've had the motherboard swapped out at the brick and mortar location where I bought it. However, the issues still persist. I guess this can safely rule out the Motherboard as the culprit here.

I still don't like the idea of thinking that 4 stick of RAM, from 2 separate purchases, are all dead.

Since swapping the motherboard, I have repeated some of the tests above. Single RAM sticks, XMP/on/off, reformatting windows, flashing BIOS etc. I have yet to actually boot into windows this time around.

I've attempted the suggestion by @helper800, however with the memory module in Slot 4, I cannot seem to boot into MemTest, it seems to get stuck on "Gathering Memory Controller" or whatever that step says. I left it for over half an hour and it never made it past this step.

I now have even less clue what to do with all these parts now, and my stress, anxiety and panic levels are reaching all time highs.

 

[jonbuder]

Since you ruled out motherboard, it sounds like the issue may lie in the CPU, since it does play a role in interacting with the memory.

I agree that if none of the RAM sticks pass the test, it's very unlikely that all 4 are bad.

Most importantly, and I know it might be super hard to accomplish, is to stay calm and not let it affect you personally to the extent it sounds like it already is. Mental health is health, and you've only got one body and one brain in this life, no need to let a hobby hold so much importance that it does harm to you.

Sure, you might lose time and maybe even money, but spent time equates to knowledge and spent money can be replaced.

 

[helper800]

So, quick update.

I've had the motherboard swapped out at the brick and mortar location where I bought it. However, the issues still persist. I guess this can safely rule out the Motherboard as the culprit here.

I still don't like the idea of thinking that 4 stick of RAM, from 2 separate purchases, are all dead.

Since swapping the motherboard, I have repeated some of the tests above. Single RAM sticks, XMP/on/off, reformatting windows, flashing BIOS etc. I have yet to actually boot into windows this time around.

I've attempted the suggestion by @helper800, however with the memory module in Slot 4, I cannot seem to boot into MemTest, it seems to get stuck on "Gathering Memory Controller" or whatever that step says. I left it for over half an hour and it never made it past this step.

I now have even less clue what to do with all these parts now, and my stress, anxiety and panic levels are reaching all time highs.

Have you built the PC outside of the case to make sure there is no short. If this is done and nothing changes I would get a loner AM4 CPU and see if that changes anything. Your PSU is from 2009 with a 7 year warranty. How old is the PSU? Do you have any way of trying a different graphics card? If you can get a loaner CPU try to get one tested that has onboard graphics so you can rule out the GPU for certainty. The CPU and motherboard are usually only diagnoses of exclusion, and cannot be assumed are the problem until all else are ruled out.

You will get through this. Take a deep breath and relax, you will have a computer up and running soon. Just be patient and go through the motions of fixing the issue.

 

[level42]

Since you ruled out motherboard, it sounds like the issue may lie in the CPU, since it does play a role in interacting with the memory.

I agree that if none of the RAM sticks pass the test, it's very unlikely that all 4 are bad.

Most importantly, and I know it might be super hard to accomplish, is to stay calm and not let it affect you personally to the extent it sounds like it already is. Mental health is health, and you've only got one body and one brain in this life, no need to let a hobby hold so much importance that it does harm to you.

Sure, you might lose time and maybe even money, but spent time equates to knowledge and spent money can be replaced.

Have you built the PC outside of the case to make sure there is no short. If this is done and nothing changes I would get a loner AM4 CPU and see if that changes anything. Your PSU is from 2009 with a 7 year warranty. How old is the PSU? Do you have any way of trying a different graphics card? If you can get a loaner CPU try to get one tested that has onboard graphics so you can rule out the GPU for certainty. The CPU and motherboard are usually only diagnoses of exclusion, and cannot be assumed are the problem until all else are ruled out.

You will get through this. Take a deep breath and relax, you will have a computer up and running soon. Just be patient and go through the motions of fixing the issue.

I appreciate the reassurance here.

I haven't technically built the PC outside the chassis, but I have rebuilt this PC 5 - 6 times now. And, I've made a very conscious effort to double check for shorts of any kind.

I've also packed up and shipped back the Kraken Cooler that I bought, since it was used from the get-go, and I've seen numerous reports that the NZXT Cam software can cause BSOD errors, and it's one variable I want to omit. I've opted for the Corsair H150i AIO cooler.

I'm dropping off the entire computer build to Memory Express tomorrow evening, and leaving the whole build in their hands. I've pretty much exhausted all I can test with the equipment and resources I have. I figured for $50 to get myself out of this nightmare, why not.

There likely wont be any updates for several days, but when I get word/the system back from Memory Express, I'll be sure to post an update, just incase someone else happens to be perusing information to troubleshoot their own issues.