High cpu temps while idling?

[Mephistopheles Faust]

Hello guys.
I have gotten myself a 5950x processor and i have noticed that despite my best efforts to decrease the temp, the cpu idles between 45-50 defeault, in a 3fan aio with half an open case. 50-65 while maxing gta out, and up to 75 solving differntial equations with very high step counts.

Are these idling temps normal? I have read online that many users of the same processor share the same experience so maybe anyone here can help ?

[_I_]

idle temps are meaningless
if the room temp is close to 30c 40-50c will be normal

you can check the cooler install, make sure it has new thermal paste

as long as its hitting and holdingits max clocks at load ts fine

[Mad Scientist]

Idling seems a bit off but without knowing the cooler, the percentage actual during "idle", any OC being done and the fan speed it would be hard to say. I have a single fan but very large heatsink, running usually 28-32C idle at 0%-5% CPU

That CPUs max temp is 90C, so the temps are overall well within range entirely.

[Mephistopheles Faust]

iCUE H150i RGB PRO XT
room temp is 20c
the thermal paste i have tried over 6, pointless to list them all
Normal crossair fans,
By idling i mean a Cpu usage between 1-4%
Under gaming load it depends on resolution and game but hovers between 10-20% usage
And under heavy load, around 50-60% with CFD, 20 million Elements of air, or under very high count Verelet to solve interactions between bouncy spheres and other physics problems. Hovers around 75c.

Have tried a 3fan radiator nzxt with the same results, so i stuck with the crossair icue since it made no difference and both are copper surfaces.

64GB 3200mhz ram and rtx3090 for gpu. Which hovers at very acceptable temps, but the cpu is wild

[r.linder]

High idle voltages and temperatures are normal since Zen2's release last year; the same complaints came then and they persist.

There's not much you can do about it because 1usmus doesn't have a power plan compatible with it yet, the windows power plans are garbage, and the CPUs are new so there needs to be more AGESA updates to optimise it.

[[☥] - CJ -]

Cooler and CPU contact might not be correct

Temp readings could be wrong if you're using multiple sources, try HWMonitor with all other temp programs closed

Case could be lacking in airflow
-

OR
Your CPU isnt downclocking which would make it run hotter than normal at idle.
If Intel Speedstep is disabled that is likely why.

[r.linder]

☥ - CJ -님이 먼저 게시:

Cooler and CPU contact might not be correct

Temp readings could be wrong if you're using multiple sources, try HWMonitor with all other temp programs closed

Case could be lacking in airflow
-

OR
Your CPU isnt downclocking which would make it run hotter than normal at idle.
If Intel Speedstep is disabled that is likely why.

Intel speedstep? It's a Ryzen 9 5950X, not an i7-5960X. Speedstep wouldn't even be available on ANY AMD processor because that's an Intel exclusive technology.

[Mad Scientist]

75C with only 50-60% CPU, now I'm curious to see what you'd hit under a full 100% load.
Gonna agree with CJ, contact is likely wrong - also hope you've been cleaning each time you've been using various pastes. Lack of airflow to the rad is also an entirely possible scenario.

I only hit 61C at 100% load with an air cooler with a bit of an OC, so.... yeah.

[Mephistopheles Faust]

Would be quite horrified if the contact was wrong. The cleaning and purifying of the surface is top notch, from paste to paste. The only thing i have ever had problems doing is unstaining copper from gallium particulates. Other than that i know how to prepare surfaces well. I do lift the cooler from time to time to check contact and thus far has been flawless. I will run more tests

[[☥] - CJ -]

sorry my mistake, not used to Ryzen sku's.
But either way, see if the clocks are going down or staying up.
AMD has its own version of Speedstep.

[Mephistopheles Faust]

I would like to give an update to this. Msi afterburner has no default cpu recording option that works. I used HwiNFO64 and integrated it in Afterburner and it shows an idle temp of just 34c. That is at least 15c more than nzxt which is currently showing 50 stable. While HWiNFO64 gives:

Cpu CCID1 (Tdie): 35c
Cpu CCID2 (Tdie): 34c
Cpu Die average: 36c
Cpu (Tctl/Tdie): 42c

But the bios gives intermediate values, i will use Ryzen Master and see what i get.

[[☥] - CJ -]

30s sounds more accurate, so i may have been right about the software being the issue
I would still use HWMonitor though

[_I_]

use hwmonitor, it logs min/max temps

if the max core temps are more than 5c apart, redo the thermal paste

[r.linder]

dcthecook님이 먼저 게시:

I would like to give an update to this. Msi afterburner has no default cpu recording option that works. I used HwiNFO64 and integrated it in Afterburner and it shows an idle temp of just 34c. That is at least 15c more than nzxt which is currently showing 50 stable. While HWiNFO64 gives:

Cpu CCID1 (Tdie): 35c
Cpu CCID2 (Tdie): 34c
Cpu Die average: 36c
Cpu (Tctl/Tdie): 42c

But the bios gives intermediate values, i will use Ryzen Master and see what i get.

Most monitors misreport things when it comes to Ryzen systems, especially HWinfo64. You're better off just using Ryzen Master to track the stats.

[Callahan420]

dcthecook님이 먼저 게시:

I do lift the cooler from time to time to check contact and thus far has been flawless.

Lifting the cooler all the time is most likely causing bad contact and air bubbles.

I hope you are not unseating and then reseating without changing the paste. Once you have installed the cooler you should leave it be unless you need to change paste or it has bad contact with cpu.

[r.linder]

dcthecook님이 먼저 게시:

Would be quite horrified if the contact was wrong. The cleaning and purifying of the surface is top notch, from paste to paste. The only thing i have ever had problems doing is unstaining copper from gallium particulates. Other than that i know how to prepare surfaces well. I do lift the cooler from time to time to check contact and thus far has been flawless. I will run more tests

As Callahan420 said, you should be replacing (wiping off all of it, cleaning the IHS with alcohol, using new paste) the paste every time you take the cooler off. If you just put it back on, you're making it worse because you're introducing air bubbles that lessen the contact between the IHS and cooling plate, which obviously reduces thermal transfer and can eventually render the cooler effectively useless if contact is repeatedly ruined like that.