New PC is spontaneously rebooting

[opethdisciple]

Built a pc for the parents last week using cobbled together spare pc parts.

Ryzen 3700x (Was in my pc for the last few years)
MSI B450M-A PRO MAX (bought new)
Corsair Vengeance LPX 16GB 3600 C18 (bought new)
Kolink Modular Power 500W 80 Plus Bronze Modular Power Supply (new)
970 Evo 500Gb ssd

I set the ram to the DOCP profile and things seemed OK. However that day after setting the pc up the computer spontaneously rebooted.

Bad sign. I decided to stress test the pc by downloading Cinebench 24 and it was OK for the 10 min multi core test and has been fine for the last two weeks.

However this evening it has spontaneously rebooted again.

My hunch is the ram. I have set it to its default speed of 2667 rather than the docp profile of 3600.

Suppose we hope for the best.

 

[wookiee87]

Kolink Modular Power 500W 80 Plus Bronze Modular Power Supply (new)

kolink psu's dont have a very good rap tbh, are you using a dedicated gpu in the build and if so what gpu are you running?
I would strongly advise looking over your psu first, you would mostly likley need a 650-750w psu depending on what gpu you have. look at something else than kolink too.

 

[Zeeflyboy]

Oddly enough I’ve just been trouble shooting a similar problem (although mine were seemingly random hard shutdowns with GPU fan ramping to max)… didn’t seem to follow any rhyme or reason, didn’t seem related to workload. In fact I never had one while doing demanding things like gaming.

Turned out at some point I had put the ram back in but not fully clicked it down… since reseating the ram again I’ve yet (fingers crossed!) to have it reoccur.

Worth checking everything is well seated (ram, power cables etc).

I think if none of the above helps, then as above PSU would be my guess - had more than my fair share of OCP reboots thanks to transient spikes from the stupid 30 series cards until I changed PSU! Do you have another one you can swap in to test?

 

[opethdisciple]

kolink psu's dont have a very good rap tbh, are you using a dedicated gpu in the build and if so what gpu are you running?
I would strongly advise looking over your psu first, you would mostly likley need a 650-750w psu depending on what gpu you have. look at something else than kolink too.

Only running a Nvidia 710.

This pc is just used as an office pc. No gaming.

For now I have turned off the docp profile on the ram hopefully that's the issue.

If it isn't then it will be the PSU.

 

[philo-sofa]

Check Windows event viewer, filter by severity and it should give you some indication of what caused the most recent reboot. Let us know what it says the cause is.

 

[Tetras]

For now I have turned off the docp profile on the ram hopefully that's the issue.

If it doesn't work you may want to try a positive curve to overvolt the CPU slightly at low load.

 

[Nasher]

Last time I had this it was GPU memory overheating. You won't see it on temperature readouts.

Another thing which can do it is an issue with the mains. Commonly happens when something with a motor is switched on nearby on the same ring (washing machine, hoover, extractor fan, etc) can cause a dip. Seen that before too, kitchen extractor fan switched on, PC crashes.

 

[opethdisciple]

Check Windows event viewer, filter by severity and it should give you some indication of what caused the most recent reboot. Let us know what it says the cause is.

Good idea. I'll have a look

 

[something daft already!!]

Have you checked for firmware updates to the Motherboard? I had a similar issue and an update resolved it for me.

 

[opethdisciple]

Have you checked for firmware updates to the Motherboard? I had a similar issue and an update resolved it for me.

Yea bios is up to date.

 

[opethdisciple]

Ok, so in the event view I see some of the following errors:

Event 18 WHEA-Logger

Reported by Processor Core
Machine Check Exception
Cache Hierachy Error

---

And another similar tothe one above but
Bus/interconnect Error

A quick google of the error and I find this.

Whats a coincidence is the ram they are talking about in these posts is a Corsair C18 kit (mine isnt rgb though)

 

[philo-sofa]

Ok, so in the event view I see some of the following errors:

Event 18 WHEA-Logger

Reported by Processor Core
Machine Check Exception
Cache Hierachy Error

---

And another similar tothe one above but
Bus/interconnect Error

A quick google of the error and I find this.

Whats a coincidence is the ram they are talking about in these posts is a Corsair C18 kit (mine isnt rgb though)

Zen 3 is incredibly fussy about memory. DOCP is not exactly perfect either.

If I were you, I'd re-enable DOCP and run memtest for a few hours to see if you can either replicate the error or get a BSOD, that would at least allow you to confirm the issue. Have you had any lock-ups since you slowed down your memory?

Other possibilities include an issue with the CPU; the 5800X does have a very complex and frankly huge cache. It's possible the thing isn't getting quite the voltage it needs for its cores or cache, but I'd hold off on testing that for now.

 

[Tetras]

Ok, so in the event view I see some of the following errors:

Event 18 WHEA-Logger

Reported by Processor Core
Machine Check Exception
Cache Hierachy Error

---

And another similar tothe one above but
Bus/interconnect Error

A quick google of the error and I find this.

Whats a coincidence is the ram they are talking about in these posts is a Corsair C18 kit (mine isnt rgb though)

Are you still getting these with XMP/DOCP disabled?

It is a pretty common error (or it was when Zen 2 was released) and it means the CPU is unstable.

Turning off C-States or increasing the minimum CPU voltage (through a positive curve) fixed it for some people, but many of these CPUs had to go back.

I know that LPX had issues with Ryzen, but afaik that was mainly the older LPX kits, not the newer "Ryzen optimised" ones. It can be tested relatively cheaply with some crappy used 4/8GB 2133/2400 stick

 

[opethdisciple]

Are you still getting these with XMP/DOCP disabled?

It is a pretty common error (or it was when Zen 2 was released) and it means the CPU is unstable.

Turning off C-States or increasing the minimum CPU voltage (through a positive curve) fixed it for some people, but many of these CPUs had to go back.

I know that LPX had issues with Ryzen, but afaik that was mainly the older LPX kits, not the newer "Ryzen optimised" ones. It can be tested relatively cheaply with some crappy used 4/8GB 2133/2400 stick

The cpu was in my gaming rig for a few years then moved to one of my workstations for another year or two and now given to my parents. Nothing wrong with the cpu.

My bet is it is the ram.

Every Corsair ram kit I have owned on Ryzen since 2017 I have always had issues with. But this stuff is meant to be optimized for Ryzen...

Corsair Vengeance LPX 16GB (1 x 16GB) DDR4 3600 (PC4-28800) C18 Optimized for AMD Ryzen - Black

I have since turned off the docp profile and hopefully that's the end of it.

However the pc is not turned on often and only has sporadic use so don't know yet if that made any difference.

I am still within the return window however I dont have the packaging any more.

​

 

[Gabby]

Every Corsair ram kit I have owned on Ryzen since 2017 I have always had issues with.

Then why do you keep buying Corsair?

 

[opethdisciple]

Then why do you keep buying Corsair?

It was cheap. I needed something cheap and cheerful.

 

[Drumroll]

AM4 ryzen had been fussy about RAM. Got better as new gen were released, but still may be the odd combination that won’t play.
From my Ryzen 1600, to my Ryzen 3900x, then Intel 12700K, the Team Group 8Pack 3200 was brilliant. Was able to run 4 sticks at their rated speed and clocks. Later tried the 3600 kit with the 12700K and worked no issues, but Intel wasn’t as picky regarding RAM.
The only time I had ever to dial back the RAM was when playing Battlefield V. Any other game, benchmark or stress test wouldn’t trigger any issue.
Another thing to check is if the correct voltage is being applied to the RAM when using XMP/DOCP profiles. Had occasions where everything else was set correctly, but the voltage.
PSU issue wouldn’t surprise me.
After going for some nice Seasonic, no coil whine, random crashes or high pitch fans to worry about.