PC power shutting off

Joined
Nov 25, 2020
Messages
30
I just built out a new rig over the weekend.

AMD 5900x
Asus Crosshair VIII Hero (wifi)
EVGA 3080 FTW3
H100i RGB Pro XT
EVGA 750W GA Supernova
4x8 GB DDR4-3600
Corsair Commander + RGB Hub
6x 120mm fans (including the ones on the AIO radiator)

Twice now I've experienced during gaming the PC power just cuts out. I've been trying to keep an eye on the temps and they don't look bad. It idles at 30-40 C and during normal gameplay when I alt-tab over it reads between 50-75 C. That said, all I'm going by are the thermals as reported by the Asus software and the Corsair iCue. The latter reports my coolant getting up to about 40C. I've had the PC on all day while I'm working; just playing music, browsing, or with a TouTube video up. No issues. All that is mostly idle temps, with an occasional increase into the 40s. Coolant is 29-33C

The AIO pump is running at 2300rpm ("balanced" setting).

Since there's no log being written anywhere that I know of that tells me why the power is getting cut, I need some guidance on further diagnosis. My first thought is cooling, but the CPU temps looks OK. Maybe the motherboard temps? PSU?

Thanks.
 
Last edited:
I've added a report from OpenHardwareMonitor for just idle temps. Temperature sensor #4 and #5 jumped out at me. I can't imagine those are the CPU temps. I can't find any data on what the 5 sensors correspond to.

UPDATE: I also added the output from HWMonitor, which looks more like the Asus output and doesn't seem to have values for the sensors reporting the 100+ C temps.
 

Attachments

  • OpenHardwareMonitor.Report.txt
    92.5 KB · Views: 0
  • HWMonitor.txt
    116 KB · Views: 0
Last edited:
750W should be more than enough for that setup.

I also have a Corsair Commander+RGB Hub and 6 120mm fans. 32GB DDR-3600 RAM. I'll add this to the original post as well.
 
Yes, this is becoming the "new normal" when pairing high core count CPUs with powerful GPUs. When all of the cores start up at the same time, within milliseconds of each other, the instantaneous power draw is quite high which can trigger some PSU power protections (OCP, etc.) and cause a full and immediate system shutdown. In most cases it is just down to the design of the PSU and what it was designed to handle in terms of instantaneous power draw. In other rare cases it exposes weaknesses or flaws in the PSU's design. It's also hard to predict which PSUs have the problem unless it has been found by others and posted online (such as in your case).

I ran into this problem on my system (see sig). I originally specced an 850W PSU, but after experiencing sudden, repeatable, instantaneous power shutdowns from two separate but identical brand new 850W units, I "upgraded" to the 1000W model and the problem totally disappeared. I don't need 1000W, but the 1000W model has a higher OCP limit compared to the 850W model, apparently, so it fixed the issue.
 
OK, got off the phone with EVGA. The tech acknowledged the issue and they are sending me a different PSU. Yes, this happens with some regularity to a small set of users, but not 100% of the time. So no recall. The tech seemed pretty confident that 750W should do me. I needed to put up the MSRP of my current PSU as collateral. Upon receiving the new PSU, as long as I send them back my current PSU promptly, the collateral comes back to me. That's pretty good customer service, I have to say. There really needs to be a warning message about this at POS though.
 
It'll just end up being another spec we'll have to look out for. Like cold cranking amps on a car battery. It's not just total capacity, but tpd as well.
 
This has popped up before, like when Haswell spec psus became a thing.
Earliest I recall was back when I ran a pair of 6800ultras in SLI I had to overbuy a psu for that rig.

Sitting on an EVGA 1200 P2 that's powered like 6 diff builds z270-x299-multie Ryzen builds-current sig
Easier to overbuy and not worry about incremental draw issues.
 
Back
Top