cancel
Showing results for 
Search instead for 
Did you mean: 

Processors

Iskaldt
Adept I

AMD Ryzen 9 5900x spikes to extreme temperatures when launching games/heavy apps.(100-112 degrees)

Hi there!

My first time on this forum, but i'm experiencing some overheating issues that i cant figure out so thought i would try here.

PC specs:
AMD Ryzen 9 5900x
MSI Radeon 6900xt
MSI MPG X570 Gaming Pro carbon
Corsair vengeance 3600mhz ddr4 (4x8gb)
Evga supernova 1300 G2 (Evga 750w since last night)
MSI mag core liquid 360R (Cooler Master MasterAir MA620M OLD cooler)
Corsair Force series mp600 1TB. x2
PC bought in May 2020. And have ran fine up until a month ish ago with the same setup.

XMP has been on since new
"Game boost" on since new
Power mode on performance in windows since new.
Have not overclocked anything manually and never touched any volt settings.

My cpu temperatures have randomly started spiking to extreme temperatures when launching games and heavy apps. But its not consistant, or atleast, it wasnt..
I have had no issues up untill about a month ago.
I launched Hunt showdown. Just as the game was starting up, my fans ramped up to 100% instantly and pc shut down.
I figured this was from room beeing very warm at the time so didnt think much more of it.
Once it had restarted, i would launch the game again and it was just fine.
I figured it was time to change the paste and clean out the cpu cooler. so i did. But the problem was still there.
Since my cpu cooler was a closed one where i couldnt change the fan (Cooler Master MasterAir MA620M), i swapped it out for a new MSI mag core liquid 360R AIO.
Pc ran fine for a while, but then it happened again when launching The Finals.
Same thing again, restart pc, launch game again just fine.

But now its gradually gotten worse and worse, and its like a 50/50 when i launch something if my pc will shut down or not. I was in the middle of a pvp match in FF14 last night and out of nowhere my pc shut down, this is the first time it has happened in the middle of gaming.
i have watched the temperatures while playing games and they are usually around 60-65 degrees area on both GPU and CPU.
Since this has been so inconsistent, but now getting worse and worse, i have no idea what is causing it.

The temperatures spike to anywhere between 100-112 degrees just as i launch something(readout from AMD adrenaline software and Ryzen master), and then steady's out to 60-65 area IF i get past the initial startup.

I have changed the PSU to a new temporary 750w Evga to test if it was my Psu sending nasty power or dying, but the same problem is still there.
Brand new AIO cpu cooler, re-seated and applyed paste an additional time to make sure it wasnt bad contact, but no luck.
I launched 3D Mark stresstest last night just to see what happened, and everytime just as it loads up the temperatures instantly spike to 100-112 area for a second and then either drop down, or my pc shuts down. seems to be a 50/50. If i get past the loadup, the temperatures steady to 60-65 degrees area during the whole stresstest.

I used to have custom fancurves, But changed them to dafault, and have tried with all fans running max speed. still same issue.

I Cleared Cmos and nothing changed. Disabled "game boost" in the bios and my PC can run games now, but its now in the high 80 to low 90 degree area when it ramps up.
Ran cinebench after disableing Game boost and it was generally in the high 80's area on all cores and around 70 on single core test.
Starting cinebench with game boost profile on in bios would kill my pc before i even got to start the test, as just the startup of the program blasted my pc to 112 degrees and boom.

Also clean installed my windows just incase it was software related but no change.


Hoping someone here might have an idea what is causing this. as it makes no sense to me since the PC has run just fine with no issues for years and suddenly this started.

Thanks.

0 Likes
1 Solution

Thank you for a very detailed reply! But i have solved the problem today.
The issue seems to have been as i initially expected. The power supply. But there was a second culprit.. i did not have all 3 radiator fans connected to the same header, causing them to lag behind on ramping up when needed. So only 1 of the fans were actually reacting fast enough to the extreme temperature spikes,while the others took their time. therefore the temperatures outran the cooling power.

I came to this conclusion today, by connecting all 3 fans to the same cpu header using a split. Put my bios setting back to "game boost" and ran cinebench multicore test, max temperature hit was 82 degrees.

I then unplugged my Pc and put my Evga 1300w back in, and my pc shut down in the middle of cinebench test.

Plugged the 750w back in again, did the same test again. and now everything runs fine and temperatures are normal. i also ran 3d mark gpu,cpu and stress tests and no issues, temperatures look good again.

As for this happening the first time with the air cooler, i think the cooler itself was simply not good enough anymore. maybe clogged with dust or something and couldnt keep up. then the PSU started its slow dying ontop of temperatures beeing high, causing a confusing situation.. if memory serves the temperatures hit 100-105 degrees that i had observed.

So basically,

PSU starts slow dying, cooler going bad and cant keep up with temperatures.
Install new cooler with bad fan configuration, cooling still cant keep up.
PSU gets worse and worse, causing pc to shut down more often every day under load. meanwhile temperatures are extreme.
Installed new PSU, but fan configuration was still the same, so temps were still high.
concluded that maybe it wasnt PSU, installed Old one again, PC starts crashing under load ontop of high temps.
Changed fan config. temperatures were more normal, but pc still shut down under load.
Changed PSU once again. this time with new fan config ontop. No crashes anymore, and temperatures normal.

View solution in original post

13 Replies
misterj
Big Boss

Iskaldt , probably a bad application displaying temperature. The maximum temperature for the 5900X is 90C, so I doubt you are really seeing 112C. Please run Cinebench R24 and post a screenshot of Ryzen Master (RM). Thanks and enjoy, John

0 Likes

I Have ran both ryzen master and Adrenaline next to eachother. They both give out the same readout.
I cant post a screenshot from Ryzen master because my PC shuts down, wich i beleave i have written in my post.  But heres a link of one of a handfull of my recordings. This is me starting cinebench. The APP not the test. i dont even get to the test part before the pc shuts down. This same thing happens when i launch games. So i dont think maximum temperature is 90 works.

https://www.youtube.com/watch?v=UMwDnAZSCzU

0 Likes

Thanks, Iskaldt. Please do not run but one utility at the same time, and make that RM until we sort things out. It looks like the 90C limit works as your PC shuts down quickly. Unfortunately it was too slow at reducing speed and voltages to throttle the processor short of shutdown. Something is bad wrong with your processor cooling. When an application is launched, Windows issues tasks and Ryzen goes to work as fast as possible, so we should expect spikes at the start. Since you have changed so many things, I suspect a power problem to the pump. When the pump stops, temperature goes up quickly. I run my pump and all fans on 12 Volts directly from the power supply. This may be too noisy for you. Please get an adapter cable (I use SATA to fan) and plug the pump into it. Plug another 3 or 4-wire fan into where the CPU was plugged so your MB thinks you have a running CPU fan. Google has many complains about your cooler clogging up. Please check it out. Running the pump at full speed should help. Have you RMAed the cooler? Thanks and enjoy, John.

0 Likes

The cooler is new, it hasnt clogged. and the pump is already running at full speed.
This same issue happens when i have a fat cpu aircooler attached aswell.
I havent changed anything other than the cooler. And my pc has worked fine with the same setup untouched for almost 4 years. the only thing i have done to it is clean dust and re-apply new thermal paste. But once this started happening out of nowhere i swapped the cooler to a 360 AIO and it hasnt helped. These spikes happens so fast that the system doesnt seem to be able to react fast enough. i have even experianced it NOT crashing after hitting 112 degrees.

0 Likes

Thanks, Iskaldt. if you decline to run the test I suggest, I cannot help you further. AMD Warranty Services. John.

0 Likes

yeah this is clearly a waste of time anyways. im not interested in repeating myself over and over.
I have already told you that the pump is running at full speed already.  ill figure this out somehow.

0 Likes

Here is a video of the night i was running stresstest on 3d mark. i was abit slow to press record but you can see at the start of the video the temperatures spikes to 111.1

and it still runs. https://www.youtube.com/watch?v=X40hksEMpKA

0 Likes

Thanks, Iskaldt. You should at least get the pump speed displayed and see what happens when the temperature spikes. John.

0 Likes

You might have an intermittent AIO MSI CPU Cooler. Many Users in the past have had issues of overheating using MSI MAG AIO CPU Coolers. In fact MSI had a recall last year concerning their AIO Coolers: https://www.msi.com/Landing/liquid-cooler-swap-application

MSI MAG CORELIQUID 240R / 360R Product Announcement

Dear MSI users,

Recently, we have received feedback from some of our users that they have experienced a drop in heat dissipation efficiency when using the MAG CORELIQUID 240R / MAG CORELIQUID 360R liquid coolers.

Our preliminary investigation has identified that a small portion of the liquid coolers produce sediment that can cause a blockage. This, however, will not cause any damage to your system since the processors are equipped with a protection mechanism against overheating.

MSI cares deeply about our users’ overall experience and we simplified our processes to provide faster product replacement services.

Users can identify whether they have an affected products by looking up the product serial number here. We will provide a comprehensive product replacement (MAG CORELIQUID 240R V2 / 360R V2) service to all affected products that have abnormal heat dissipation ( Example: CPU temperature is higher than 60°C while idle and the cooler is properly seated with thermal paste applied between the CPU and cooler ).

The problem could be that you have a partial clogged AIO system.

Best way to eliminate the AIO is by using a cheap powerful AIR CPU Cooler to see if it overheats.

As mentioned, here are your specs for your processor:

Max. Operating Temperature (Tjmax)
90°C
 
Default TDP
105W
 
Recommended Cooler
Liquid cooler recommended for optimal performance
 
90c is not the maximum shut down temperature for the processor but rather Maximum Operating Temperature that the processor is engineered to run safely without damage for long periods. Once it reaches 90c it automatically starts to throttle to cool down the temperature to 90c or below. If the temperature keeps rising it will eventually reach its critical shut down temperature and  shut off the PC before the processor or PC gets damaged.
 
I personally suggest you open a MSI Warranty ticket to see if your AIO Mag CPU Cooler needs to be checked or replaced.
 
EDIT: Read you other replies. So with both AIO CPU Cooler the CPU spikes to 112c, Correct?
 
Either you have a defective processor, AIO, or Motherboard sensors.
 
Try opening a AMD SUPPORT - Warranty ticket to see if they believe the issue could be due to a defective processor from here: https://www.amd.com/en/forms/contact-us/support.html
 
They will notify you if you need to open a AMD WARRANTY TICKET.
 
NOTE: AMD WARRANTY for Processor in Box Official Retail Box is 3 years and your are the original owner with a valid Retail receipt.
 
 
0 Likes

Thank you for a very detailed reply! But i have solved the problem today.
The issue seems to have been as i initially expected. The power supply. But there was a second culprit.. i did not have all 3 radiator fans connected to the same header, causing them to lag behind on ramping up when needed. So only 1 of the fans were actually reacting fast enough to the extreme temperature spikes,while the others took their time. therefore the temperatures outran the cooling power.

I came to this conclusion today, by connecting all 3 fans to the same cpu header using a split. Put my bios setting back to "game boost" and ran cinebench multicore test, max temperature hit was 82 degrees.

I then unplugged my Pc and put my Evga 1300w back in, and my pc shut down in the middle of cinebench test.

Plugged the 750w back in again, did the same test again. and now everything runs fine and temperatures are normal. i also ran 3d mark gpu,cpu and stress tests and no issues, temperatures look good again.

As for this happening the first time with the air cooler, i think the cooler itself was simply not good enough anymore. maybe clogged with dust or something and couldnt keep up. then the PSU started its slow dying ontop of temperatures beeing high, causing a confusing situation.. if memory serves the temperatures hit 100-105 degrees that i had observed.

So basically,

PSU starts slow dying, cooler going bad and cant keep up with temperatures.
Install new cooler with bad fan configuration, cooling still cant keep up.
PSU gets worse and worse, causing pc to shut down more often every day under load. meanwhile temperatures are extreme.
Installed new PSU, but fan configuration was still the same, so temps were still high.
concluded that maybe it wasnt PSU, installed Old one again, PC starts crashing under load ontop of high temps.
Changed fan config. temperatures were more normal, but pc still shut down under load.
Changed PSU once again. this time with new fan config ontop. No crashes anymore, and temperatures normal.

Great troubleshooting.

Doesn't your AIO come with just one Radiator cable for all 3 fans?  Yes that would and could cause your processor to overheat if only 1 of 3 Radiator fans was running at full speed when it was overheating.

I have read that PSU can cause a CPU to overheat by supplying bad voltages but your issue is one I haven't heard before that a bad PSU was causing the CPU temperature spike high enough to shut down your PC.

I read that EVGA PSU are not that great compared to other PSUs.

Thanks for the update and new troubleshooting tip for the future.

NOTE: Is the EVGA PSU under Warranty still?

Honestly i dont remember seeing any extra cables in the box for the fans. and they dont have a daisy chain. so all in all its 4 cables. ive had the cooler for a while just laying on the shelf and i had already yoinked some fans from the box before actually using it. so i might have put the cables somewhere i dont remember, if there even was any.
I havent heard about cases like this before either. tho i have heard about psu's delivering bad power before. im just glad its all sorted now atleast. And no, i bought the psu used from a friend so i dont have any warranty. im unsure of the age of it but im suspecting around 5 years ish..

Hi elstaci. I have edited my Solution response to be more factual. i was laying awake in bed thinking about this and I have realized that the PSU did infact not directly affect the temperatures of the cpu. but due to how this has all went down led me to beleave it did. i realized that the temperatures were infact most likley fixed during cinebench test with 1300w. but it still shut down due to PSU beeing bad. i was away from my pc when it shut down so i made assumptions that it was due to high temperatures.. Anyways, it was bugging me, and it have fixed it now.