• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

7800x3D randomly hits 90C, readings make no sense?

Joined
Mar 1, 2021
Messages
122 (0.08/day)
Processor R7 7800X3D
Motherboard MSI B650 Tomahawk
Memory 2x32GB 6000CL30
Video Card(s) RTX 4080 Super
Case Lian Li O11 Air Mini
Power Supply Corsair RM1000x
My CPU runs somewhat warm, probably since im using a mediocre cooler (Arctic 34 esports Duo). During gaming it pulls 40 watts and is around 70 to low 80 C which im fine with. But sometimes apparently it also pulls like under 50W and goes to max 90C?? First screenshot is an example of regular gaming load, 26% usage 41W and 75C, all fine. (I was using 140fps cap)

20241015042447_1.jpg


In this second screenshot i open up ingame letter, which should be easy to render right? GPU wattage drops by 90% as expected, CPU load drops to 15% 31W and even 100MHz lower clock speed. But somehow temps go up to 90C which makes no sense? Same happens when i open the map. And it isnt just MSI afterburner reading, i can hear the CPU fan spool up. Anyone know why does this happen? Mafia: DE is the only game where i've encountered this. Could these be faulty readings by the motherboard? Would a bios update fix it?


20241015042034_1.jpg
 
I had a whole post written-out, and realized that while it may be an errant reading, I've experienced similar on a similarly-designed aircooler (HDT).

Heatpipe Direct Touch/Contact CPU coolers perform weird* and terrible on MCM SoC CPUs.
IIRC, I had a FSP Windale 6 (or very-similar HDT cooler) on my R5 3600 and R5 5600. It was just barely sufficient.
When I moved up to a 5800X3D, the cooler became wildly insufficient, and would inconsistently thermal throttle while gaming (like you're seeing).

Your current HSF is great for 8000G APUs on AM5, and most-all of Intel's desktop offerings.
-Just, not AMD MCM CPUs, and especially-not AMD X3D MCM CPUs.

*Basically what's happening is the IHS makes for a lacking 'coldplate'; only a select-few heatpipes are getting the majority heatload from the CCD or IOD.
(In the worst-case) when that/those heatpipe(s) gets thermally saturated, it/they stop working and whatever was in best-contact with it starts to overheat.


I'd recommend something like the Thermalright Phantom Spirit, if you're on a budget and wanting to stick w/ aircooling. (~$30-40 USD)
1744325532408.png

 
Last edited:
Sounds like your GPU fps limit is being ignored or something. The temps would follow work done imo. Good luck finding a better cooler.

The Noctua coolers are very good.

We're all having hardware issues with X670E. imo

:toast:
 
Sounds like your GPU fps limit is being ignored or something. The temps would follow work done imo. Good luck finding a better cooler.

The Noctua coolers are very good.

We're all having hardware issues with X670E. imo

:toast:
Zero issues with my x670e for the entire time I've had it.
 
Zero issues with my x670e for the entire time I've had it.
Do you have a Ryzen 9000 based system. I had no issues with 7900X3D until the last BIOS and a CPU swap to 9800X3D.

This CPU and BIOS combination is a real problem. You know im going to have to try switching out the CPU to see if that solves it:toast:
 
Go into your bios and try a negative offset of -20 in PBO. It should in theory lower temps a little when gaming.

Although a Thermalright Phantom Spirit or Peerless Assassin might be the way to go
 
Last edited:

Better dissipation will mean better performance and not better temperatures. Check this out, it explains in detail how your processor works
 
Last edited:
Nothing here either, think it's called core optimizer set it to -20 as already said and should be good. Just make sure the system is 100% stable just in case crashes start if they do try -15.
 
My CPU runs somewhat warm, probably since im using a mediocre cooler (Arctic 34 esports Duo). During gaming it pulls 40 watts and is around 70 to low 80 C which im fine with. But sometimes apparently it also pulls like under 50W and goes to max 90C?? First screenshot is an example of regular gaming load, 26% usage 41W and 75C, all fine. (I was using 140fps cap)

View attachment 394488

In this second screenshot i open up ingame letter, which should be easy to render right? GPU wattage drops by 90% as expected, CPU load drops to 15% 31W and even 100MHz lower clock speed. But somehow temps go up to 90C which makes no sense? Same happens when i open the map. And it isnt just MSI afterburner reading, i can hear the CPU fan spool up. Anyone know why does this happen? Mafia: DE is the only game where i've encountered this. Could these be faulty readings by the motherboard? Would a bios update fix it?


View attachment 394489
A screenshot of HWiNFO sensors mode would be helpful when this 90C for no apparent reason is happening. A full screen shot would be perfect so it can show as many sensors as possible.
 
Back
Top