Okay thanks. Have installed it now. Will try run it tonight or tomorrow. Will let you know the results once I’m done with it.
Memtest86+ takes hours to complete. I suggest you run it over night.
1 Like
Yeah, that’s exactly what I was planning to do.
am I the only one expecting another chilling overnight murder? 
3 Likes
The plot thickens…
See you all tomorrow for the complete, tense, and terryfying show-down!
1 Like
I did install zsh and run mprime, same result, the torture test was running for about 4-5 miniutes until i killed it …
1 Like
I ran the test over night. It said Tests passed: 48/48 (100%)
0 Errors as well.
I’d say it’s not the RAM.
hmm… the killer thinks they are hiding their tracks well by misdirecting us, but the causal loop is tightening and so the mistakes are bound to happen soon.
Sit tight folks… also keep a book or file of holy mprimes nearby, incase they get cheeky to visit you.
I’m just warning you that memtest86+ does not detect some errors accurately:
- DMA transfer under certain circumstances.
-
Probability of bit error is low like random (One time test is not enough, you need repeat test more than 10 times.) It is related to:
- Temperature
or - Humidity
or - EMI
or - Voltage stability
or - Some external factor
- Temperature
I suggest you to test every single RAM with mprime. Do not test all RAMs together.
You would easily find which one is faulty. I did that by myself.
What is the CPU temperature saying when you run “mprime -t”? May be you have a thermal issue.
If every single RAM has the same result “mprime got killed”, then most likely your CPU or motherboard is faulty.
I will test it tonight when I come back home. Thanks for the advice.
Will test it as well when I come back home. However I have a 360mm AIO, so I doubt that the temperature will be an issue as I’ve never had thermal issues with my current setup. But will test just in case.
I got Ryzen 9 5900x (PBO2 auto, Curve Optimizer per core setup), 32GB (2x16GB) DDR4 G Skill Ripjaws CL16 3600MHz, and RX 6800XT.
Not sure if that matters, but just letting you know.
I also have a Ryzen 9 5900X. When I played around with curve optimization and undervolting I used mprime to test stability of the settings. When the settings were to aggressive mprime got killed after several runs.
I suggest you first test the factory defaults with mprime.
1 Like
I really hope that that is the case for me. Everything else sounds like I need to replace some part of the PC.
Will try to reset Curve Optimizer to default settings (disabled), and will see if that fixes the mprime homicide
After I disabled Curve Optimizer, mprime wasn’t murdered.
I am guessing that some CO values were unstable on some cores and caused therefore mprime to be murdered 
Thank you everyone for helping me out.
@keybreak murder suspect was found and apprehended!!!
2 Likes
We can sleep safe…For now!
3 Likes
Oh well, I was happy a bit too soon. It seems that the real perpetrator wasn’t caught after all.
Even with Curve Optimizer disabled now, it becomes murdered.
Sometimes it runs for a min, and gets murdered, sometimes it gets murdered as soon as I start it.
I am so confused.
mprime -t gets also murdered very quickly.
That is not so good. But it is good that you tested with mprime.
As a next step I would follow the proposal from @Zesko : Take out all RAM and test with only one RAM module at a time. With that you either identify the one RAM module that is broken or you identify that your motherboard has an issue.
By the way, which CPU temp do you have during testing?
Could it not also be related to driver issues and/or kernel modes, or chipset-issues on linux, respectively, with such a new and highly powered computer, they seem to be using?
At least, that’s what I was thinking at first, before pondering any HW failure involved, and not hoping that.
