Idle BSODs on 3700x

[ReD-EyeD]

There is also this. I'm not sure how reliable this info is, though.

 

[Shou Miko]

When RAM fails in a 4 pass test haven't you considered to contact Kingston support/rma about this problem?

Even if your RAM is on the QVL list doesn't always make it work out of the box not said Asus haven't tested a similar kit.

Even if you just activate A.M.P. it doesn't ensure that your ram is stable or sitting the speed manually because everything over the default supported memory speed is an overclock.

 

[ReD-EyeD]

When RAM fails in a 4 pass test haven't you considered to contact Kingston support/rma about this problem?

Even if your RAM is on the QVL list doesn't always make it work out of the box not said Asus haven't tested a similar kit.

Even if you just activate A.M.P. it doesn't ensure that your ram is stable or sitting the speed manually because everything over the default supported memory speed is an overclock.

It failed on the 5th pass and only once out of 3 times. I haven't contacted anyone yet as I'm still trying to pinpoint the problem. It's pretty hard to rma it with this lockdown stuff and I need a computer right now even if it works unstable.

Yeah, it's not on QVL list.

I did test the ram on 2133 and it still bsoded the same although kinda rarer, but I'm not sure as bsods are fairly random in nature, but always in idle/low load and never on heavy load as I can pass RealBench stress test for hours.

 

[Shou Miko]

You still say that you have bsod doing idle so there is something wrong.

The ram I use isn't on the QVL list for my board but they are working like a claim no issues at all this proofs that ram can work even they ain't on the QVL list but there might be issues.

 

[bug]

So first stick bsoded 2 times with DPC_WATCHDOG_VIOLATION and SYSTEM_SERVICE_EXCEPTION blaming ntoskrnl.exe and amdppm.sys

Trying out the second one

So, did the second stick crash as well?

 

[ReD-EyeD]

So, did the second stick crash as well?

Yes, but it went through all 8 memtest passes. Didn't memtest the first stick tho as it bsoded aswell.

 

[bug]

Yes, but it went through all 8 memtest passes. Didn't memtest the first stick tho as it bsoded aswell.

Bummer. Best case scenario, the RAM needs RMA. Otherwise, you're looking at components that just don't work together (CPU+RAM+mobo) and will need replacing. Sorry.

 

[spectatorx]

Bummer. Best case scenario, the RAM needs RMA. Otherwise, you're looking at components that just don't work together (CPU+RAM+mobo) and will need replacing. Sorry.

That is what it is and that's what i had on my 2700x. I was unable to make this setup work stable under literally ANY conditions. It was freezing and bsoding for no reason even after replacing components to new units. OP just has to face that and alter hardware choices, there is no way to fix this situation.

To be honest i'm surprised there is no talk about it in media as such cases are not rarity and are limited to all series of ryzen, to whole am4 platform and incredibly hard to analyze and prove. Just once you fall into setup of cpu+mobo+ram giving you such issues you have to alter your choices by replacing all 3 components to different models, replacing with new units will do nothing.

 

[ReD-EyeD]

@bug @spectatorx
What do I troubleshoot first then? RAM? Mobo?

 

[eidairaman1]

Dude replace the ram, it's failed tests. Stop going in circles and do something!

 

[ReD-EyeD]

Dude replace the ram, it's failed tests. Stop going in circles and do something!

Which one then? There are so many opinions on that.

Also can memtest86+ errors indicate some kind of cpu problem or it's strictly ram test errors?

There is also this: https://community.amd.com/thread/244029?messageTarget=all&start=125&mode=comments
A lot of people with very similar problem are blaming (and fixing) the cpu itself.

 

[spectatorx]

Which one then? There are so many opinions on that.

Also can memtest86+ errors indicate some kind of cpu problem or it's strictly ram test errors?

There is also this: https://community.amd.com/thread/244029?messageTarget=all&start=125&mode=comments
A lot of people with very similar problem are blaming (and fixing) the cpu itself.

Ryzens have built-in memory controller so yes, ramtest if shows any errors it is either problem with ram or cpu. I will repeat myself again: in my case i replaced both to new units and problems were still there.

 

[bug]

@bug @spectatorx
What do I troubleshoot first then? RAM? Mobo?

Right now you're looking at RAM failures, so I'd start with that. Zen2 loves DDR4 3600 anyway.
Though, as @spectatorx said, it could be a faulty memory controller inside the CPU.

I guess the best thing would be to borrow some RAM sticks and try to see which one it is, but I don't know if you can do that atm.

 

[ReD-EyeD]

I guess the best thing would be to borrow some RAM sticks and try to see which one it is, but I don;t know if you can do that atm.

Sadly, I can't.
For now I upped the voltage on ram to 1.4 to see how it goes. Should I run memtest overnight like that?

 

[bug]

Sadly, I can't.
For now I upped the voltage on ram to 1.4 to see how it goes. Should I run memtest overnight like that?

Don't bother. Memtest isn't bulletproof, even if it passes you may still get the BSOD. And if overvolting kills it, you're also out of warranty.

 

[Nabarun]

Don't assume it's too simple a solution! CLEAN your PC! Take out the GPU and RAM and thoroughly clean using an air duster and then with 99% isopropyl alcohol. Clean the fans (gpu and cabinet), connectors, slots and traces in mobo. This has cured a lot of my nightmares multiple times, including very recently.

 

[bug]

Don't assume it's too simple a solution! CLEAN your PC! Take out the GPU and RAM and thoroughly clean using an air duster and then with 99% isopropyl alcohol. Clean the fans (gpu and cabinet), connectors, slots and traces in mobo. This has cured a lot of my nightmares multiple times, including very recently.

It's too recent a build to have accumulated any significant amount of dust. (But yes, I fixed a laptop once by cleaning its CPU cooler.)
First time I'm hearing about cleaning traces, though. How do you do that?

 

[ReD-EyeD]

So I managed to solve a little part. I can now run prime95 with all workers running with an offset to CPU voltage for -0.0125v
Bsods still occur tho, but at least this little thing gives me faith.

Overnight memtest86+ showed no errors in 5 passes

 

[bug]

So I managed to solve a little part. I can now run prime95 with all workers running with an offset to CPU voltage for -0.0125v
Bsods still occur tho, but at least this little thing gives me faith.

Overnight memtest86+ showed no errors in 5 passes

I'm not sure what faith you get from that, it seems to indicate the problem is the memory controller and you'll need a new CPU.

 

[ReD-EyeD]

I'm not sure what faith you get from that, it seems to indicate the problem is the memory controller and you'll need a new CPU.

This is exactly where my faith comes from - pinpointing the problem after so much headache. There is also this thing with prime95 failing to load 2 threads (1 core). I just hope that ALL the problems will be solved with a new cpu and not just this particular thing.

 

[bug]

This is exactly where my faith comes from - pinpointing the problem after so much headache. There is also this thing with prime95 failing to load 2 threads (1 core). I just hope that ALL the problems will be solved with a new cpu and not just this particular thing.

Ah, ok. Well if getting a new CPU is not a hassle, then you're getting close. Best of luck.

 

[ReD-EyeD]

Ah, ok. Well if getting a new CPU is not a hassle, then you're getting close. Best of luck.

It actually is, especially in these quarantine times, but at least it's a solution. Have to buy a new one while I'm dealing with the faulty one return.

 

[Aquinus]

I did ran memtest86+ from an USB as I mentioned before. Two times it went fine for 4 passes and another time there were errors on the 5th one.

http://imgur.com/lZo5Dx5

I'm also concerned about prime95 instantly turning off 1-2 cores before them reaching 100% load


I have the latest bios already

Hey folks. I know that I'm late to the party, but that memory address seems really low. I don't know exactly how memory addressing works on modern X86 processors or in the context memtest86, but even on microcontrollers, these are usually logical addresses, not physical ones in actual memory. If that's the case, then 2cc90 is only about 183k in, which could easily place the failing memory addresses in cache if the cache is addressable.

Are we certain that the CPU is stable?

Edit: I should have finished reading the thread.

 

[ReD-EyeD]

Are we certain that the CPU is stable?

Edit: I should have finished reading the thread.

I very much appreciate the answer! Do you think that it's a cpu problem as well?

 

[ReD-EyeD]

Update:
So, I'm on a new 3600x and I have yet to encounter any problems here. If there is anyone with a similar problem with a decently old gen3 ryzen you should really look into changing it as apparently earlier chips have this problem.