CPU temps issue
CPU temps issue
I have encountered some temperature concerns with the following components:
i7-13700K
MSI MPG Z790 CARBON WIFI
32GB Corsair Vengeance RGB DDR5-6000 36-36-36-76
RTX 2070 Super Gaming X Trio
Phanteks EvolveX with Corsair QL140 (3 front, 1back, 2 top)
Noctua NHD15
This summer I first observed elevated temperatures reaching up to 100°C on certain cores. During gaming sessions in my room, I reached temperatures as high as 35°C. I attempted to lower them by modifying the fan curves and adjusting the case fan settings to match CPU temperatures, assuming that internal heat buildup might be the cause. Even after these adjustments, the issue persisted. I then reoriented the cooler for better airflow direction and refreshed the thermal paste during the process.
Current idle temperatures range from 30-36°C across all cores and 35-40°C on the package. Stress testing with CPUZ (P <230W) resulted in P cores reaching 88-100°C, E cores 80-85°C, and the package 98-100°C. The temperature jump to 100°C typically occurs within 1-2 seconds at the start of the stress test. Given these high readings, it’s unclear whether the recent adjustments made a noticeable difference—I’m uncertain about the exact throttling behavior before the change.
Another concern involves the game Squad. Upon launching it, usage and temperatures spike to 100°C on affected cores. In some instances, the PC freezes or shuts down entirely, requiring a forced shutdown to restart. This issue first appeared during summer after an update, and it continues even after two major updates. I’m unsure if the update increased startup CPU load or if the crash stems from temperature issues or a new bug. Recently, I played GTA5 again; the temperatures were between 50-60°C, and the PC crashed twice while playing. This suggests another potential problem unrelated to CPU temperatures.
In summary, regardless of the Squad issue, are there significant differences (>10°C) between cores, and do the overall temperatures seem reasonable considering the thermal paste application? Are the sudden spikes from idle to 100°C within 2 seconds during startup or testing normal? Could the temperature-related problems be linked to CPU performance?
MSI MPG Z790 CARBON WIFI BIOS information
What BIOS version are you running on your motherboard?
The cooler was rotated to improve airflow direction, and the thermal paste was refreshed during the process.
Did using the NT-H1 tube included with the cooler affect your thermal paste application?
One recommended action for your case is to remove the front fascia, including the dust filter, to check if temperatures have improved.
Running the newest Bios (no beta) version 7D89v1H. I recently purchased some new duronaut paste, so the tube shouldn’t be old. If I recall correctly, opening the front didn’t noticeably affect temperatures, but opening the side did help a bit under partial load. During the stress test it still throttled up a bit even with panels open. All that is typical before the changes. Now, opening any panels doesn’t make much difference, so airflow should be improved overall.
Display images of your bios configurations. Focus on CPU and OC sections, though fan settings and others can matter too. Default options aren't always what they seem—they might force your cores to work harder. Reaching 100 could happen because the BIOS interval for checking temperatures is sufficient for the CPU to hit high temps.
These are all the CPU-related configurations.
Fan Ramp up is configured at 0.1s.
I have disabled the MSI Game Mode, which means the remaining settings follow the default MSI configuration. However, based on your guidance, I reviewed the power limits—there are 253 sustained watts, which is a bit higher than the default. The processor appears to reduce its output from that maximum to about 200W gradually, maintaining the target of around 100C throughout the test.
Please note that the system keeps the 100C at least on one core for the entire duration of the stress test, not just briefly.
Reference: https://imgur.com/a/DRn7P24
Your CPU is experiencing slowdowns because it's undergoing thermal throttling. Power usage drops when a CPU has to slow down due to heat limits.
A cooler isn't sufficient to keep a processor running at 253W or higher. To minimize thermal throttling, lowering the long-term turbo power cap to 200W or lower can help. This signals the CPU to begin power limit adjustments, potentially reducing thermal throttling. The impact on overall performance is likely minimal. Throttling is simply a way to manage heat.
To achieve peak performance from an Intel 13th or 14th Gen CPU, an AIO is often necessary.
13700K, 13900K, 14700K, 14900K.
An update on my case: with an -0,1V offset I achieved stable CPU performance at 5.3GHz, P-core average around 90°C and spikes reaching 95°C, P-cores 5°C cooler than the rest. The temperature didn't appear to be a major issue for crashes, though I'm relieved I've optimized them now. The high voltage spikes (and resulting heat) seem to have damaged the memory controller, preventing it from handling my kit at XMP settings (as reported with many 13700K units). Currently, I only receive unstable OC messages after re-flashing the same BIOS version. Now I'm running RAM at 0.2GHz without any problems so far.