Performance issues reported with i9 14900K under heat stress. Cinebench results indicate problems.
Performance issues reported with i9 14900K under heat stress. Cinebench results indicate problems.
I recently assembled my PC a few days ago (the specifications are listed below) and managed to get it running smoothly with Windows 11. Over the past day, I ran a Cinebench multi-core test and achieved a score of 1900. The online benchmark suggested a maximum of 40,000 points. I also noticed the CPU temperature reached around 100°C. What might be going wrong? Checked the BIOS—everything seemed to be installed correctly, including the fans and the AIO unit. All connections appear to be properly made. Any suggestions would be greatly appreciated.
Gigabyte Aorus 4090 master 4 x 32GB DDR5 RAM i9 14900K Corsair H150i AIO ASUS RogStrix Z790-A Wi-Fi II
Are you running the single-core benchmark and confusing it with the multi-core version? These processors tend to overheat, Corsair AIOS is often average for cooling performance. They appear good but lag behind typical AIOs in heat management and are expensive. What is your fan speed curve? Claiming smooth operation doesn’t provide much clarity. Do you have screenshots or visuals to support this? Why did you opt for 128GB of RAM? Is there a specific need for that amount? XMP settings are likely disabled or incompatible with the 4x32 dimmer configuration, which could negatively impact performance. Have you adjusted your voltage settings? Overall, it seems you performed the single-core test and misinterpreted it as multi-core. If the multi-core results are accurate, there appears to be significant thermal throttling. Which monitoring tool are you using to track CPU performance and temperatures?
I'm using a specific Cinebench release, as each update affects the results. Version R23 typically ranges from 36,000 to 41,000 based on your thermal throttling and power settings, while the latest 2024 version falls between roughly 2,000 and 2,300 again under similar conditions. I'd estimate it's 2024; a score near 1,900 suggests possible issues like poor airflow or heavy background tasks, but it could also reflect modest hardware limitations. A single-core run might land in the 2,200–2,300 range, though that would require significant throttling to reach such low scores, which is unusual unless the system crashes during games.
I’m confident I performed a multi-core test. My system is used for deep learning, training large models, and running tasks like stable diffusion and other heavy computations. Unfortunately, I can’t share fan speed graphs at the moment, but I’ll try to clarify what they mean. From 0°C to 50°C fans run at 30%, then from 50°C to 55°C they reach 50%, 55°C to 60°C they hit 60%, and above that it stays steady at 70%. Regarding Corsair AIOs, there are some options that might offer better performance. I also use software like iCUE and hwmonitor, which reported around 100°C during Cinebench. As for your RAM, removing one stick or two could help if you’re experiencing issues—let me know if you’d like advice on that. Thank you for your support!
Maybe because you mentioned running it at 70% fan speed. If this were different, yes—but with Cinebench it's very rare. 4x32GB DDR5 isn't great, yet Cinebench often ignores RAM speed entirely, so even if you hit around 3600-4400MT/s manually, it won't significantly affect the score.
Would you consider a few options for a suitable cooler? You currently have qx120 fans and a H150i, so replacing them might affect your setup. For your NZXT H9, an all-around fan configuration is possible with an NZXT AIO featuring integrated fans.
I’d increase the fan speed on your AIO initially to check if that boosts the score and confirms the issue lies with the unit itself. If it does, the Arctic Liquid Freezer III 360 would be a strong choice; the Freezer II is still decent but has seen better sales recently. Regarding appearance, I’m not confident in this area and should leave that decision to someone who truly pays attention.
NZXT AIOs offer solid options, especially for a balanced three-fan setup. They provide reliable performance and good airflow management.
Usually these are too costly for typical use. They're merely new versions of Asetek models, similar to the Corsair unit you have.