• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Is my ryzen 3600x faulty?

Associate
Joined
18 Aug 2013
Posts
225
Hi
First off sorry to the people in the taichi thread who are sick at the sight of me by now!
I've been going through the process trying to diagnose a 4 month long crashing problem with games, borderlands 3 gets a UE4 fatal error and either CTD's or locks up the pc, forza horizon 4 just closes with no error message but event id 1000 shows up in the viewer. This can happen about 20 mins into gameplay or hours.
I've RMA'd one set of memory over it which was sent to manufacturer for repair/replacement, it was 8 pack DDR4 3600 C16. It crashed less at lower clock speeds, when I ran memtest it failed on 3rd pass. I'm starting to think this is not the problem as i'm currently using a cheap but brand new set of corsair vengeance 3200mhz, it's been absolutely faultless for about three weeks but over the last 2 days the exact same problem has come in again.
I've tried all of the windows and driver fixes for all components including chipset drivers, system file checker and the BIOS is updated on the motherboard (X570 taichi). I don't think windows is the problem since it fails memtest. Running memory at exact XMP settings and timings. I tried upping the ram voltage too.
Could I have a faulty IMC which I think is on the CPU?
The rest of the system is superflower 750W 80+ gold psu, tried 2 graphics cards my old GTX 1080 and now on 3070, an asus xonar 5.1 sound card which has the software installed, CPU is cooled by a corsair H100X. I am also using phanteks braided cable extensions on the CPU.
 
using 2 slots, the first and third. I clocked the 8 pack ram all the way down to 3000 and it crashed still, i'll try with this set i suppose. I haven't actually put this set through memtest yet either. I also have a brand new set of 8 pack ram to go in, surely if I test that too and all 3 have the problem I can eliminate the possibility of ram problems?
 
using 2 slots, the first and third. I clocked the 8 pack ram all the way down to 3000 and it crashed still, i'll try with this set i suppose. I haven't actually put this set through memtest yet either. I also have a brand new set of 8 pack ram to go in, surely if I test that too and all 3 have the problem I can eliminate the possibility of ram problems?
The ram should be in the 2nd and 4th slot mate
A2 B2 this is probably your issue
 
From your manual. https://download.asrock.com/Manual/X570 Taichi.pdf

You CPU is 'Matisse'

Put your ram in A2 and B2, A1 and B1 only support 2933Mhz.

fOOAtmt.png
 
Also probably not the issue but if your running the Asus Xonar, try disabling the onboard sound via the bios, more than likely the Ram Slots though
 
HAd same issue and partucularly in UE4 engines with that same 8 PACK RAM. Swapped gfx cards from 1080ti to 3080. Same issue.

Swapped MOBO from ASRock phantom gaming 4 as it was a cheap MOBO to an MSI tomahawk - still same problem.

Swapped out RAM to Corsair vengeance 3600Mhz never crashed since. That 8 PACK RAM is fickle stuff and doesn't work with XMP 2.0 as advertised. I tried hours with adjusting volatges, SoC, entyering timings manually, DRAM calc etc. Might be OK for a bench run but for me my set wasn't stable enough for hours of gaming.

In 30 years of gaming I've never had an issue with RAM.

I see you are using the MOBO 8 PACK recommended also. I think what he recommends is for benching so only needs to be stable for duration of benches.

Was a WIN64 SHipping error thrown up in the UE4 crash reporter? Need more deets on the crashes.

But I lasted a year with my crashes and it was only really in UE4 engine. Wont be buying anything from a recommend from OCUK staff again. Need to send this RAM back and move on my old x570 MOBO that was probably fine too. 30 years and never had to mess about with RAM as much as this before. REally annoyed. But others use it fine!

Feel for ya!
 
So, thanks for that tip but it still crashes in the second and fourth slots I'm afraid! I'll try to get this image of the ue4 error uploaded

edit: easier for me to type it out, imageshack isn't what it was!
Title:the UE4-oakgame has crashed and will close

Fatal error!

Unhandled exception:EXCEPTION_ACCESS_VIOLATION writing address 0xcffffffffffffffcf

0x0000000143642633 Borderlands3.exe!UnknownFunction[]
0x0000000143643b16 Borderlands3.exe!UnknownFunction[]
and so on
further down: 0x00007ffce7c97034 KERNEL32.DLL!unknownFunction[]
W2bCWMf
 
Last edited:
Have a look in your motherboard BIOS to see what voltage the SOC is being set at. Make sure PBO is switched off. Also make sure any motherboard auto performance boost profiles are disabled. Then check and see the SOC voltage. It could be the BIOS is setting funky SOC voltages which can affect RAM stability.
 
I had SOC turned off, PBO turned off and i've clocked ram down from 3600 to 3400 so far. It's crashed twice in two weeks, one was literally as I opened forza horizon 4 on the starting video, the other one which just happened was 90 mins into playing borderlands 3
edit- i've also added more fans as part of a planned upgrade, 3 noctua p12 1700rpm mounted as intake at the bottom which has dropped gpu temps by a couple of degrees. That obviously hasn't had enough of an effect on stability. and a bit of misinformation on my part, the corsair vengeance which i'm using temporarily is rated 3600mhz so I started at that and so far because it's crashed so infrequently, i've only tried down to 3400
 
Last edited:
I had pretty much the same problem with a 3800x but it only crashed in apex legends memory related error passed every test i threw at it, could get around the crash by using ryzen master to set a fixed overclock with a set voltage. Sent it back and bought a 3800xt which doesnt crash still uncertain if it was the cpu or the motherboard having a slight stuttering problem with the new cpu
 
Two weeks on not a single crash with - 50mhz clock speed, that's definitely stable in my eyes. I might take the plunge and try that new set of 8 pack 3600mhz ch16 which I've got from before, see how it handles the tighter timings coming from the ch18 vengeance. I'd probably best compare 3dmark physics scores between them too
 
1.1V on the cpu is a bit low
With low voltage you might get performance regression
go for 1.2V
You can set the SOC voltage to 1.15V or 1.17V
The cpu memory controler runs off the soc voltage
Do not go over 1.2V for 24/7
 
It took three weeks but it crashed again! That's with CPU at 1.15V, -50mhz, ram at 3600 with IF at 1800, SOC at 1.17. At this rate i think i'm prepared to just live with it if it's only going to crash that often, just think only about 17 more crashes till the cpu gets changed anyway! But knowing my luck it will gradually worsen again
 
Back
Top Bottom