Issue with cooling system NOCTUA NH-D15 + 13700k
Issue with cooling system NOCTUA NH-D15 + 13700k
It doesn't work as expected; the CPU shouldn't reduce performance when using a Noctua D15 cooler. My 10900k, which is also power-hungry, stays below 90°C even after overclocking at 5GHz with manual voltages on Prime95 SmallFFTs (maximum stress). During auto-voltage overclock, it hit 90-91°C in similar tests on Prime95 SFFTs. I haven’t checked all the details—just that I’ve been running it stock since installation. The dual-tower D15 setup with two 14cm fans covers most of the heat well. I also tried a single tower Noctua U14, which didn’t overheat under stock conditions. This isn’t typical for laptops; it suggests an issue might be thermal paste, a faulty cooler, or a bent IHS requiring replacement. In extreme cases, a CPU delamination could be necessary.
Hello, thank you for your response. I applied a contact frame and observed an increase in power draw to 240W after testing the CPU with CinerBench R23. Previously, with the motherboard's mounting frame it stayed at around 230W. The temperature rise wasn't significant, though it still reached 100°C. Using Noctua NH-1 thermal paste seems to be the better option to try. Regarding the cooler, check if it's faulty by monitoring performance under load—fans reach up to 1500 RPM without visible damage. Also, note that the front two NF-A14 PWM fans only spin at 1000 RPM during stress tests, even though I set them to increase to 100% once the CPU hits 70°C.
I just completed another round of disassembling and reassembling plus applying fresh paste on the cooler. My patience is wearing thin as I notice it being used on other systems. Another observation: when I remove the cooler, the thermal paste appears on the sides, leaving the center nearly dry. This doesn’t make sense to me. I’m puzzled about what’s going wrong, even though the contact frame was intended to solve it. I suspect the middle area is getting poor contact where it should be most, which could be causing overheating. I don’t know how to fix this. Here are my core temperatures after a 10-minute Cinebench R23 test. Also, I noticed that HWMonitor isn’t displaying real data for some reason. In this run, I scored around 30293 points with fans at full speed, and the CPU was drawing up to 241W. For gaming, the highest I’ve seen is a spike to 87°C in FIFA 23; otherwise, it’s usually between 50-75°C depending on the game or scene. I’m becoming increasingly frustrated because I’ve spent two weeks buying components and now I’m left with no clear solution—especially since the contact points still seem to be problematic. PS: I’m trying to prevent the deli from forming. Also, I’m wondering if my heat pipes are losing efficiency after using this cooler for a while.
I tested the Cinebench R23 myself; it looks like I surpass the 90°C mark, though my highest recorded is 95°C. The photo was taken at the 9:50 minute after completing a previous 10-minute cycle. My power draw stays consistent at 244W, so energy use matches. What differs is the score: after ten minutes I achieved 15872, which is far from your expected double. I’m sending this to Intel—those new CPUs are impressive. My processor didn’t slow down until it reached 95°C on both core sets. The temperature difference comes from the same cooler design; the bent/convex IHS affects pressure distribution. Without the cooler, the paste doesn’t spread evenly because the center gets the strongest pressure, causing spillage on the sides. When you install the cooler properly, pressure is more balanced, reducing paste leakage. I think the frame isn’t doing a great job fixing this issue. If possible, don’t force the cooler too much—avoid putting excessive pressure on the center. You might want to try again with slightly looser screws and less force. Since the peak temperature only occurs during stress tests and throttling happens only at full load, you should be okay in normal use, especially if you don’t push all cores (100%). For gaming, a max of 87°C is likely achievable on the core that hits 100°C in Cinebench, which is probably safe long-term (even if it reaches 90). If it doesn’t throttle at 100°C, you won’t lose performance. I’ve used this CPU since 2020 for gaming and never faced issues—either stability or overheating. Just enjoy the performance and ignore the numbers.