CPU temperature sensor freezes on MSI X870E Tomahawk WIFI

I

ilya.mfw12e402b7

New member
Joined
Jul 4, 2025
Messages
6
Hi, I'm another owner of a motherboard with a stuck sensor

I will not write the details, as everything has already been described on this forum and Reddit:

System:
MSI X870E TOMAHAWK WIFI
AMD 9950x3D
Arctic Liquid Freezer III 420
PCCooler YS1200
GeForce RTX 5070 Ti

If you have solved the problem with the sensor, please share the solution in this topic
 
Last edited:
Sort by date Sort by votes
New response from support team (08.Mar.2026):
Please download and update the BIOS in the link below to check if it helps. Thank you !

Best Regards,
Markus
 
Last edited:
Upvote 0 Downvote
I still need to clarify a few things, for example which motherboard the BIOS is actually for, for the non-E or E version, before I can post further information.
 
Upvote 0 Downvote
Latest resaponse from MSI Support (09.Mar.2026):
It is the latest BIOS for MAG X870 TOMAHAWK WIFI, you can test it.
However, we don't suggest you share it now and we suggest you first check if your former issue can be fixed. Thank you !
 
Upvote 0 Downvote
Add me to the list as well. MAG X870 TOMAHAWK WIFI, from the Newegg hustle so I have xmp/non-expo Corsair ram as well.
I did a lot of initial benchmarking and the temps stayed fine. Then when I tried to play with curve optimizer, I had a case where Cinebench 23 froze the temps on me.
Ive since undone the Curve Optimizer offset and tried other stress tests and hit it with OCCT once, and not again even after 4 more hours of avx512.
However tried Aida64's Stress CPU, FPU and Cache, the sensor froze only 1 min and 40 seconds in. I rebooted, reran it and sensors froze at 1 minute in.
When I have time Ill try and toggle between cpu, fpu and cache and see which one triggers it fastest. Also had it happen once in Borderlands4.
Ive only had it for 2 weeks. Newegg sadly says any returns have to be all items in the bundle, so Id have to deal with swapping the mobo, cpu and ram again. I'd rather not need to play that game. If a Reboot can reset the sensors, I dont see why they can have some code or app to run periodically or on command to reset the sensors so at least we dont need to revert to a whole reboot.
 
Upvote 0 Downvote
mrksbk said:
Latest resaponse from MSI Support (09.Mar.2026):
Click to expand...
1773303838936.png

is it the same BIOS from MSI site?
 
Upvote 0 Downvote
Was asking mrksbk to clarify :LOL: , not sure if MSI has some sort of secret BIOS other than the one on website now.
Does the BIOS fix anything?
 
Upvote 0 Downvote
i never tested opening msi center, but those weird reports of hwinfo only happens when the bug occurs ( it takes ages to load and when it does, it shows -200c and no other sensors like that ) still trying some msi orientations, will get in touch soon
 
Upvote 0 Downvote
@rafaelg.mott159402e2
HWiNFO-version is quite outdated (and also CPU-Z ;) ), the latest beta version as of today is 8.45-5945 (direct download-link).
Check if there are any improvements made on how it interacts with the sensor of your board (or if MSI can't fix it even contact the author on their forums, so a workaround to cull those extreme readings might be implemented).

When you say takes ages to load, do you mean only during booting or generally in the OS too everything feels sluggish?
 
Upvote -1 Downvote
WakeX said:
@rafaelg.mott159402e2
HWiNFO-version is quite outdated (and also CPU-Z ;) ), the latest beta version as of today is 8.45-5945 (direct download-link).
Check if there are any improvements made on how it interacts with the sensor of your board (or if MSI can't fix it even contact the author on their forums, so a workaround to cull those extreme readings might be implemented).

When you say takes ages to load, do you mean only during booting or generally in the OS too everything feels sluggish?
Click to expand...
You completely missed the point.
This is not a HWiNFO or CPU-Z issue. Those tools are just reading data — they are not the source of the problem.
What’s actually happening is a telemetry failure at firmware level:
CPU sensors (Tctl/Tdie, CCD1) freeze or disappear
IOD hotspot reports invalid values (e.g. -200°C)
Power and current readings stop updating or vanish
Ryzen Master crashes on launch
The system itself keeps running normally
That combination makes it very clear this is not a software monitoring issue, but a breakdown in sensor reporting (likely SMU/firmware related).
I’ve already tested:
Multiple BIOS versions (including MSI-provided test BIOS) currently testing some new ones that they sent me

Different monitoring tools and versions
Same behavior every time.
Also, this issue is much easier to trigger under AVX-heavy workloads (like shader compilation), but it can also happen at idle — which again points to firmware instability, not a tool bug.
So no, updating HWiNFO won’t fix a problem where the motherboard/CPU stops reporting valid telemetry in the first place
 
Upvote 0 Downvote
WakeX said:
@rafaelg.mott159402e2
HWiNFO-version is quite outdated (and also CPU-Z ;) ), the latest beta version as of today is 8.45-5945 (direct download-link).
Check if there are any improvements made on how it interacts with the sensor of your board (or if MSI can't fix it even contact the author on their forums, so a workaround to cull those extreme readings might be implemented).

When you say takes ages to load, do you mean only during booting or generally in the OS too everything feels sluggish?
Click to expand...
I dont think its an issue with the monitoring software as I have had this happen during Cinebench, Aida64 (usually within 5 minutes), OCCT stress tests. And in real world its glitched on me playing Borderlands 4 and Diablo 4.

When it happens, OCCT and Aida64 both report CPU package 0w. I neglected to check what HWinfo64 says. Also the cpu temp will be frozen in place. If it glitched at 83c, then it will stay at 83c. I know this isnt just a visual glitch in the reporting tools as the cpu fans will ramp up and stay at the speeds thinking its still at 83c. If temp were to go down, ehh the fans are still running a little extra. not the end of the world. However what if my temps spike again to 95c like it does when compiling shaders, then its stuck at a lower speed.

I reached out to MSI about an RMA, and sadly they dont offer an advance replacement, and Im also responsible for the shipping to them. What a horrible position to be put in. And im outside of the retailers return window since I thought I might be able to solve it myself. Now I'm tempted to buy another of the same motherboard and use that to ship the faulty one, fingers crossed the new one works then I can return the newer one. Although due to this experience, Im almost tempted to bite the loss and just get a totally different non MSI motherboard, and then take my chances trying to sell this one.
 
Upvote 2 Downvote
TheGlow said:
I dont think its an issue with the monitoring software as I have had this happen during Cinebench, Aida64 (usually within 5 minutes), OCCT stress tests. And in real world its glitched on me playing Borderlands 4 and Diablo 4.

When it happens, OCCT and Aida64 both report CPU package 0w. I neglected to check what HWinfo64 says. Also the cpu temp will be frozen in place. If it glitched at 83c, then it will stay at 83c. I know this isnt just a visual glitch in the reporting tools as the cpu fans will ramp up and stay at the speeds thinking its still at 83c. If temp were to go down, ehh the fans are still running a little extra. not the end of the world. However what if my temps spike again to 95c like it does when compiling shaders, then its stuck at a lower speed.

I reached out to MSI about an RMA, and sadly they dont offer an advance replacement, and Im also responsible for the shipping to them. What a horrible position to be put in. And im outside of the retailers return window since I thought I might be able to solve it myself. Now I'm tempted to buy another of the same motherboard and use that to ship the faulty one, fingers crossed the new one works then I can return the newer one. Although due to this experience, Im almost tempted to bite the loss and just get a totally different non MSI motherboard, and then take my chances trying to sell this one.
Click to expand...
Your comment is on point, this has nothing to do with monitoring software, since it will break on every single one of them, ryzen master, occt, afterburner, any

Im currently testing some custom bios they are sending me, no success so far, if anything solves the issue, i will post it here

Your board was not on warranty? im trying to fix this but i have the option to send it to my local msi representation and get a refund, but im trying to avoid the massive headache, plus if they manage to fix this with bios updates, will help everybody
 
Upvote 1 Downvote
Back
Top