Extreme temps on Intel I7 14700KF

OjoMeister

New member
Joined
Oct 20, 2025
Messages
2
Hello everyone,

Here is my system:
CPU : Intel Core i7 14700KF with Thermal Grizzly LGA1700 Contact Frame​
GPU : MSI GeForce RTX™ 4070 Ti SUPER VENTUS 3X OC​
MBO : MSI MAG Z790 TOMAHAWK WIFI​
AIO : Lian Li Galahad II Trinity 360 black​
RAM : ADATA XPG LANCER RGB 6400MHz cl32 64GB (4X16 GB @6200MHz)​
SSD : ADATA XPG GAMMIX S70 2TB​
HDD : Seagate Barracuda 2TB​
PSU : be quiet! Pure Power 12 M, 850W​
CASE : Montech AIR 903 Max​
OS: Win 10 Pro x64​
I ran out of ideas so you guys are my last resort. For the past couple of days I have been trying to figure out the overheating of my I7 14700KF processor, but whatever I do I can't seem to sort this out. My train of thoughts might jump from one idea to another so forgive me and please bear with me. I know 13th/14th gen had issues with the microcode and the power limits, for a while i ran with the stock bios settings but switched to the Intel recommended settings as soon as i could, within a week or so, but i don't know whether my CPU was affected by this or not. I have not seen any major issues with it, i mainly play CS2 and EFT, where i get plenty of fps i would say and office tasks are no issue with this build.

Since i've built this PC i noticed that the temps were a little to high but everyone on the net said that 13th&14th gen ran hot so i just thought that it was somewhat normal. At the moment when i was commissioning the PC i ran some stress tests and temps were reaching high 90C if i recall correctly, but at that time i was running 2 sticks of RAM. Since then i've "upgraded" to 4 sticks of RAM, 32GB was a bit on the low side considering Tarkov's high need of memory and the fact that i was also keeping Discord and other chome based apps which take lots of RAM space. The setup was not stable out of the box when enabling the XMP profile, 6400MHz is just too much to ask for, so i settled for 6200MHz but also had to do some voltage tweeking in order to run stable at this freq. Anyhow everything was fine the temps were still a bit high while gaming (85C-95C) but never had any issues. I remember I used Intel Extreme Tuning Utility for the CPU power limits, i'm not sure i did the bios tweeking but i probably did. In order to use the tool, VBS has to be disabled and it was, until i decided i wanted to go back to FACEIT, where in order to use their anticheat you have to enable VBS, which of course made the tool unusable and so the power limits not imposed in the system. A few mounths have passed without me noticing anything, no real issues, there were issues but manly due to poor optimisations of the games i've played, until a few days ago a friend of mine bought Tarkov and had issues running it, his CPU usage was 30/40% but had low FPS so i wanted to check how my system was doing and that's when i noticed i was almost always running at 95-100C which freaked me out. I closed everyting and opened MSI Afterburner and guess what... 70C while idleing on desktop... I panicked, with no clue as to why is this happening. I work in engineering so i imediatly started to rule out the causes,
  1. I thought about bad install of the contact frame, tried to get a torque wrench/screwdriver but no available tool is capable of measuring 6cNm specified by Thermal Grizzly. I have reseated the CPU and frame 2 times, screwing by holding the tool with 2 fingers until you cant torqe it anymore than adding 90 degrees as Der8auer mentioned. This did not fix it
I checked the AIO Pump it seems to be working. The fan speed and pump speed are liked by wire, i was going to disconnect them and connect them separately to the MBO but i din't get the chance, but still this would not really affect the temps just the control. They all run constantly at 3200RPM, I measured the temperature with a IR Thermometer, while CPU at full load (90-100C) the output of the pup is sitting at 40C with the input at 30C with arround 26/27C ambient temp. This did not fix it.​
I checked the thermal paste aplication, it was ok, reaplied thermal paste, used only Thermal Grizzly Krionaut. This did not fix it.​
I cheked the temps on the MBO with the IR sensor, it was sitting at 50C, even when measuring directly the contact frame i never saw more than 50C.​
There is no discussion about ariflow, i have a total of 7 fans in the case and now i'm running with no side panels and no filters and the temps are still high.​
I am really out of ideeas. If you guys think of something please let me know. I also attached some pictures with the system and a screenshot from HWInfo. Looking forward to your answear and thank you for your time.​
1784145338079.png
1784146868476.jpeg
1784146884185.jpeg
1784146951427.jpeg
1784146996366.jpeg
 
I checked the thermal paste aplication, it was ok, reaplied thermal paste, used only Thermal Grizzly Krionaut. This did not fix it.
Looking at that thermal paste, I would say it is not OK as it appears to be quite thick in some areas. A good application, when you remove it should look even and transparent

At this stage, I would try a different cooler, as the Coldplate may be warped or the CPU has an uneven or badly made Integrated Heat Spread
 
know 13th/14th gen had issues with the microcode and the power limits, for a while i ran with the stock bios settings but switched to the Intel recommended settings as soon as i could, within a week or so, but i don't know whether my CPU was affected by this or not.

Every true 13th/14th gen CPU (like any -K/-KF one) will be affected by the buggy microcode included in old BIOS versions. So the most important thing is not Intel recommended settings for things like the power limits, the most important thing is to use a recent enough BIOS, ideally from 2026.

You see, the Intel recommended settings are not the real solution to any degradation. When this issue first cropped up and people got instability in certain games, which stemmed from the first CPUs already being degraded, Intel had no idea what caused it. But as they scrambled to come up with some kind of solution, their initial idea was to clamp down more on the board makers, which historically allowed maxed out power limits and almost no safeguards for the CPU, in the search for higher performance in comparison to their competition. Because with all other hardware being the same, the board won't really impact the performance too much, unless one board allows higher power draw for the CPU, so that's what most of them did. So Intel for the first time handed out stricter recommendations, which were applied or at least recommended in the new BIOS versions that followed.

But even with much more sane defaults, like 253W/253W/307A - which should be the most that should reasonably be allowed for any CPU - the CPUs kept being degraded. So they later found out it was all because of voltage spikes, which chipped away at microscopic structures inside the CPU and could render it unstable over time. So step by step, they worked on updated microcode, which was supplied to the board makers and bundled in later BIOS updates. This is the real fix for the voltage spikes, that's why having a new enough BIOS version is the most important thing.

Now, on top of that, even on the latest BIOS version, a CPU like the 14700KF presents a whole different issue, in that those kind of CPU models were pushed way too far from factory, with excessively high power draw and resulting temperatures. So here, the power limits come into play, which should ideally be adapted to the cooling capabilities of the individual system. Also, most CPUs can be undervolted while remaining fully stable. Both steps I explain in the Guide: How to set good power limits in the BIOS and reduce the CPU power draw.

Now, considering all this and looking at your post and photos.

The screenshot doesn't state your BIOS version, but let's just assume you updated the BIOS and you are on a recent version. The cooling capability, with a nice 360mm AIO, and airflow-oriented case with a mesh front, and plenty of case fans, should be around the 253W power limits that are set. Certainly somewhere between 200-250W, if not slightly above.

But if we look at "CPU Package Power" (power draw of the CPU only), we are seeing a mere 130W, and it's running into thermal throttling, while your system, drive and GPU temperatures are all very low. This points to something being seriously wrong with your CPU cooling, be it a failed pump, clogged cooling circuit, or a problem with the mounting / contact to the CPU. This CPU cooler currently performs worse than a small- to medium-sized single-fan tower cooler. The CPU needs to throttle itself at only 130W, so either the heat doesn't get properly transferred into the pump unit on the CPU, or the heat doesn't get properly transported to the radiator. Even >40°C in the BIOS is too high. With your cooling, you should even see <30°C there. So it has all to do with the CPU cooler here somehow.
 

That doesn't look good.

I checked the thermal paste aplication, it was ok, reaplied thermal paste, used only Thermal Grizzly Krionaut. This did not fix it.

I would switch to Honeywell PTM 7950
but 1st fix this:


The contact frame you use is too possible to contribute to the overheating issue.
Try re-seat the contact frame, ensure it's mounted correctly.
[see details here: https://forum-en.msi.com/index.php?...z790-a-wifi-13700k.411481/page-2#post-2329751 ]
If needs, try remove the contact frame and test it without it.

P.S:
Similar to my suspicions I just found here:


I know 13th/14th gen had issues with the microcode and the power limits,

My recommendation is to use >>Forum UEFI Shell Flash & Patch Tool<<
to install latest BIOS 7D91vHI.zip for your board,
to update to the very latest uCode 0x136 (the best version so far from all),
[after loading the BIOS in the tool, Enable the uCode upgrade from Tool's Option]
and very latest ME-FW 16.1.42.2872 (auto installed along with BIOS)
_
 
Last edited:
Back
Top