i7-13700k is overheating - Need assistance with undervolting? [updated]
i7-13700k is overheating - Need assistance with undervolting? [updated]
It's possible to undervolt with XTU. It helps without strictly limiting performance, which is beneficial, though it does reduce efficiency. Through undervolting, you get a few advantages. Mainly, lower voltage means less heat, which results in reduced fan speeds on coolers, increased boost speeds, and better performance over time. A video provides a practical reference for this approach: Undervolt I7 13700k 10-150mV, Noctua D15, Cinebench R20 R23 temperatures, how hot will it get? Windows 11 - YouTube. What kind of thermal paste are you using? There are both good and poor options available.
Others have examined my homework information and suggested it might indicate the CPU cooler liquid isn't working well. This has made me think the cooler could be faulty. I hope it's not that. I assembled this PC for a Thanksgiving break project, but it seems I may need to return it and wait about three weeks before using it again.
I think an undervolt could assist. In practice, a 13700k feels like a better version of the 12900k, which was previously recognized as a space heater. Watch: https://www.youtube.com/watch?v=4jjopjkJzxA
I believe I was only using arctic silver, but other hwinfo data suggest the issue might be different
Not overly sensitive at all. The incorrect term would be more fitting.
Power-hungry components would suit you better.
You handle various tasks well, so restricting the E-cores really cuts performance. Not a good idea.
There are several factors against your setup.
First, it's from the 13th generation, and it's commonly acknowledged that 13th gen models suffer from thermal issues because of the socket constraints. This will affect temperatures.
Second, since it's from the 13th gen, PL1 and Tau are largely obsolete; we're back to just three CPU modes—idle, base, and Turbo—where only idle and full load turbo are actually used. That means a CPU that shifts from 10w to 230-250w (depending on the motherboard) almost immediately under stress tests like Cinebench or rendering with heavy core usage.
Third, it's a 230-250w 13th gen CPU, which demands at least 350w of cooling capacity (360mm AIO) to avoid thermal throttling under full load. This depends on case temperatures, fan placement, airflow, motherboard settings, and voltage controls.
Your cooling should be adequate, but combined with all these elements, the temperatures aren't unexpected.
If you remember, these chips are built to operate around that 100-degree threshold. So it's not unusual at all—it will continue pushing the juice in until it reaches there. No cooler has managed this without removing power constraints. It simply won't happen unless you apply some extremely harsh cooling. That's why setting a power cap seems like a straightforward solution. Just limit it to 200w or even 250w, and it should cool down. Let it run without restrictions and it will keep pushing until it hits those extreme temperatures and stays there.
It was definitely unusual. As soon as I performed any CPU-heavy tasks, the system would instantly spike to 100°C. After disconnecting and reconnecting everything connected to the pump, and repeating the standoffs, my PC now stays under 80°C during cinnebench runs. I also increased the fan header voltage and switched the pump off in quiet mode in BIOS. Everything seems resolved. I’ll update the original post. Thank you all very much. With the current situation... I wouldn’t need to undervolt at all. No throttling issues have appeared since the repairs. Now I just need to find a way to speed up my premiere rendering again. Thanks so much everyone.