PC Processors

[Studynxxx]

RTX 4090, 7900x, 2x32GB DDR5 6000Mhz RAM, 4TB SSD M2, Gigabyte Aorus B560M Elite AX, Aero prism ARGB Black, Be Quiet Dark Rock Pro 4

Ever since I got my PC - 2 weeks ago - I have had the problem of the 7900x overheating constantly. By this I mean that while I game in 4K max settings, especially in new AAA games, 

the 7900x quickly gets to 95C and 5500Mhz and then stays very close to that temperature, namely around 90-95C but usually around 93-95C. The clock speed begins to fluctuate once it reaches the max temp of 95C, between 4700Mhz-5500Mhz, up and down, down and up.

All components are brand new. I made sure that the DRP4 doesn't have the plastic film on the bottom - it does not. Re-seated the DRP4 and also re-applied the thermal paste on the CPU. Went into the BIOS and put the CPU into the 105W Eco Mode, applied Curve Optimizer All Cores Negative 20, to no avail: Under heavy load -eg in Games, Cinebench, etc - it still quickly reaches 95C and stays there, rides that line. At idle I get 54C-65C

Also tried taking off the glass side panel and game for 10 minutes: The temps remained the exact same so it doesn't seem like airflow is the issue (even tho I thought so!)

Is this normal or is this still wrong? Is this expected behavior? Seems super weird to me but I might be ignorant on this.

[johnnyenglish]

Hey,

With EcoMode 105 you shouldn't be getting 95C at all!

Ive tried EcoMode on a 7950X and barely goes above 65C, but I havent tried curve optimizer yet. 

Try without curve optimizer and do a regular undervolt by offset, benchmark it to see how it went.

Its just for troubleshooting the final temps.

UPDATE your bios in the process or limit vSoC to lower than 1.3v if you have expo enabled. 

Good luck 

[slash00x]

You may have put an abundance of paste on the cpu or running on a fan for cooling? You only need just a dot of paste in the center, press and twist if possible but not so hard that might damage cpu.

I am running the 7900x and do not even get close to that, 62c in Farcry New Dawn. Using a Corsair H150 radiator. Thats at peak 5400+mhz.

[johnnyenglish]

You really have to over paste it seriously like a mad man to get adverse effects, I already troublehooted this with my colleagues at work, I put a 3x pea sized paste to the point where the paste reached the board PCB and it cooled almost normally, maybe 5ºC worse, maybe.

The only other option would be the plastic peel still in the CPU block.

[slash00x]

I put one small dab of paste in the very center of the cpu. Thats all you need. pea size is way too much and 3 that size is way overboard. Best bet is to get a radiator and apply just a small dab in the center. 

Be sure to update the UEFI with the latest. 

"Update, 4/27/2023: AMD says that it has identified the issue that was causing some Ryzen 7000 CPUs to burn out and that it has released a new version of AGESA (an AMD-controlled part of every Ryzen system's BIOS) "that puts measures in place on certain power rails on AM5 motherboards to prevent the CPU from operating beyond its specification limits, including a cap on SOC voltage at 1.3V."

Full load for me with a corsair radiator with a number of other software's running in the background max 170c.

[slash00x]

This may help I'm using Arctic Silver 5.

[seatommyboy]

First off, the mobo listed in you post (Gigabyte Aorus B560M Elite AX) appears to be an intel board. Assuming this is a typo, I would guess your board is actually a B650 chipset? There are a few different versions of the B650M Elite AX as well. but I am going to assume you have rev 1.2. To caveat off @slash00x, I have read a few articles that this is a bios (AGESA) issue. I would recommend updating your bios and lower CPU core voltage to around 1.15-1.2. On a dual boot system, I keep my cpu voltage at 1.1 in bios (linux reasons) and use msi center to set it to 1.15 when booted in windows. In linux, my peak temps under full load=~71C, in windows (using occt)=~62C.

[dariuszpawlak85]

I also built my set on GIGABYTE B650 GAMING X AX, rev, 1.0 + RYZEN 9 7900x

I have been struggling with computer stability problems from the very beginning. It hangs at various times. temperature approx. 90 C, with AiO cooling. on Saturday I received the ASUS TUF GAMING X670E-PLUS WIFI motherboard, during the replacement I noticed 3 black pins on the bottom of the processor, as if they were burned out? I have no problem with heating on the new motherboard, the computer has been working stably for 2 days, it doesn't hang up, the screen doesn't turn off, it works as it should. The only thing I noticed is that when I only have a motherboard and a processor, artifacts appear on the screen, but there is no problem after inserting the RTX. I think my processor is damaged by the GIGABYTE motherboard

[seatommyboy]

Y'all should check this out.

AMD Ryzen 7000 Burning Out: EXPO and SoC Voltages to Blame (AMD Responds) 

[dariuszpawlak85]

As for me, I didn't overclock anything more than selecting EXPO mode in the bios. Do you think that what you see in the photo is a burnt processor?

[Farcuen]

A burnt processo do not turn on your CPU, do not worry, and the issue caused about the burning out is already fixed it.

In any case just to be sure, flash your BIOS with the latest provided by your Motherboard producer.

Also be carefully handling CPU.

If you need help I am here.

Let me know

Kind regards

Farcuen

[Farcuen]

High temperature usually can be categorized in two type of issue:

• Software.
• Hardware.

First and foremost access to BIOS, and check the temperature into it before OS are launched.

Important note: Before to check the temperature through the BIOS properly please restore the default settings.

If temperature are high when you are into BIOS an still consistent around 95C° that's mean you have an hardware issue somewhere, like bad thermal paste application or a non good cooling solution for your CPU.

Usually into BIOS temperature is a little bit higher, but not that much as using demanding application or bugged software which can cause a constant overloading of the CPU.

Theoretically for that beast of CPU I expect something between 45C or 80C with some spikes randomly occurs by normal software, browser, YouTube videos and etc.

If on the BIOS the temperature still around 95C restore the BIOS settings to the default, if the issue persist you have to bought an AIO, to cooldown that beast, also because your AIR Cooler is to weak to cooling down that.

At this point you can easily do one thing to understand if there is some software which cause that issue:

Look the task manager which you can open with CTRL + SHIFT + ESC

Image text are in Italian, but you have to sort the CPU column to check which software highly overload the CPU, there is also another possibility, where you have to uninstall ARGB things and OEM software like AOURS Control Center, and Only Install driver, keep system simple with only drivers, if your temperature cooling down, some software of that cause the Issue.

I said that because, one friend of mine had a similar issue with an Intel I9 not about the temperature but with the CPU overloading and helping him the as I said to him it is possible motherboard utility cause issue, in his case the ARGB of MSI Dragon Center or related software was the guilty.

If after all of this, the issue persist can be possible you have a faulty CPU, so return it.

Personally I recommend for that type of configuration a:

-360 AIO cooling for the CPU.

- 3 Intake FAN, and 1 exhaust (the other exhaust 3 are included into the AIO cooling).

Let me know

Kind regards

Farcuen

[Angeluk]

Is there any software intensively using your CPU? Check clock speeds in task manager and bios? Wat is your cooling? Did you try reinstalling your CPU cooler? I suggest you undervolt. I keep mine at 5.1ghz and 1.065 volts, that is stable clock for 7900X, at least with my MB and PSU, but i guess it will work for you too Let us know how you doing

[Farcuen]

Undervolting it is a good tips but first and foremost a CPU should run out of the box properly, and you do not have to fix the issue with it, if the issue come from another source.

Put the scenario he has a faulty CPU with the undervolting process you can fix it, but the issue at hardware level still there (of course if he has, he must before follow what I explained into my reply).

Undervolting is good for cooling CPU when CPU works as expected, and not for bug, software issue or hardware damage.

[Angeluk]

At this stage he does not know if the CPU is communicating properly with the MB, as he previously said his CPU does not respond to bios curve optimizer... No harm trying! At least in some sense he will get a hint where the issue may be coming from. My 7900X was getting exactly the same temps out of the box, 65 at idle and 95 while gaming (not even 100% workload). I suggest you wait for his response before throwing any other guess