New build hard reboots with "email" load

[eltharynd]

SOLVED: It was a 24pin PSU cable extension from cable mod.

 

 

Hello... 

So I built a computer in early 2021 (don't hate me I was super lucky).

 

Asus prime x570 pro

5950x

3090 TUF OC

Corsair RM850x White *2018* (850W)

4x 8GB corsair dom plat rgb

 

 

the PSU was my only choice as the 1000+ Watts I ordered with the parts got delayed for multiple months then ultimately I had to chose another one (ordered in Ord, built in late Jan)

 

Obviously I overclocked the shit out of it. After a while I started noticing some hard reboots, no blue screen, PC immediately turns back on. It legitimately looks as if I pressed the reboot button.

I immediately thought it was an unstable OC, so I procedurally started to tune it down... but I've been running for a while with no OC at all now and it still does that...

 

I've been quite too busy with work to properly troubleshoot till now, but someone on the discord suggested it might actually be the PSU, some stuff about handling current spikes poorly. Thing is PSUs are still waaaay out of stock so i can't test with a proper one.

 

The thing i noticed is: it never crashes while under heavy sustained load... 8 hours gaming and streaming sessions never had a hiccup.

Reading an email in work ours, watching twitch, being idle when downloading a torrent.. that's the stuff that makes it reboot.

 

So I'm here, a humble man, asking for advice...

What do you guys think it might be?

 

Edited May 19, 2021 by eltharynd
SOLVED

[emosun]

How many ram sticks are in it

[8-Bit Ninja]

7 minutes ago, emosun said:

How many ram sticks are in it

I know that sounds like a strange question. But Zen 3 has some pretty bad stability issues with 4 sticks of RAM 

[eltharynd]

wow.. i have 4x 8GB corsair dominator plat rgb 3600... is that for real? KEKWait 

i ran memtest multiple times both with custom timings, OC, docp and just stock... never had an issue and they're now running at stock

 

 

wouldnt that result in a bluescreen? or a system log?

 

i guess it cant hurt to only use 2 for a while and see, but im somewhat doubtful

[eltharynd]

UPDATE: I just took 2 sticks out, The heatsinks where fucking hot to the touch... I was monitoring temps of sensors but seems like while dissipation for main component is enough the ambient temp of the case makes vrm and ram way hotter than i'd expect... not sure if it's just this particular config but i ramped up the fan speeds a bit just to move some more air.... if it doesn't crash anymore ill try putting the sticks back and see

[eltharynd]

UPDATE: just rebooted again... Wasn't that

I put the sticks back in and kept the fan profile cause temps were way more agreeable.

 

Something else I'm trying: at the beginning i experience some issues with a 10gbs network card that was preventing auto sleep for windows 

I figured since it was giving "power troubles" why not try and take it out... at this point i need to try everything i guess

[eltharynd]

UPDATE: just rebooted again so it wasn't that either... what do you guys think? easiest thing would be to test with a different PSU but those are unfindable now.... is there any logging for that kind of hardware failure?

 

here's the error on the even viewer altho i suppose it wont help much

crash.evtx

[Mark Kaine]

8 hours ago, eltharynd said:

Thing is PSUs are still waaaay out of stock so i can't test with a proper one.

Took me 20 seconds to find

https://www.amazon.com/gp/aw/d/B01MZ8A8NW/ref=psdcmw_1161760_t1_B008Q7HUR0

 

but i dont know if the price is normal (probably not)

Only 1 left btw

 

 

And another

https://www.amazon.com/gp/aw/d/B01N5NWKHH/ref=psdcmw_1161760_t1_B01MZ8A8NW

(cheaper)

2 left

[eltharynd]

Amazon doesn't ship where I live anymore.. and when it did stuff marked in stock often really wasn't tbh