Ridership increases in overheating issues by 14 to 16 degrees.
Ridership increases in overheating issues by 14 to 16 degrees.
Hello! Your 5800X began running warmer than expected despite your setup. After assembling the system with a Noctua NH-U12S, Arctic MX-6 paste, and a well-ventilated case, idle temps stayed low at around 27°C even at 4.375 GHz. Over time, after upgrading the motherboard BIOS, idle temperatures rose to 40-42°C. This prompted some maintenance steps like cleaning dust, adjusting fan positions, and securing screws. Although the temperature remained high during startup, it’s possible the initial build had minor inaccuracies or environmental factors affected readings. If you haven’t changed power settings in Windows, consider tweaking them slightly for better stability. Let me know if you need more details!
I also performed a Cinebench evaluation (this was my second run on the system) and observed peak performance around 63.5°C during a full multi-core load test. While it wasn’t outstanding, the score dropped significantly—900 points lower compared to the first run, even though the clock speed remained unchanged. Still curious about why that happened...
Could be that the new version came with a new set of voltage curves or standard settings for PBO. I don't know enough about this to say anything with confidence, but that would be my first guess to look into. If this is indeed the case, it should be documented or described in some OC forum or somewhere. So maybe look for the exact BIOS versions and changes. Also, if you still remember which version you had previously (or have a good idea) you could try to flash it back to that version and see if the temps go down again.
So 27C idle might look low even with a custom loop enabled. Perhaps the old BIOS contained an issue and you've got incorrect sensor readings. Your current idle and load temperatures are all good, same goes for your load temps. BIOS updates could change performance settings and overclocking options, possibly leading to a lower score. Also remember that extended overclocking might wear out the CPU over time, causing subtle performance drops that could cost you a high score in a benchmark.
It makes sense if the sensor had an issue capturing data accurately. The clock adjustment seems reasonable given the 6% decline over seven months. Thanks to everyone for your thoughtful feedback before I pushed too hard.
Hi Ashton, I haven’t found a solution yet, but I’m looking into this thread because I plan to assemble a PC with the 5800X soon (including air cooling and a properly ventilated case). Have you thought about how temperature shifts might just reflect seasonal climate changes? For instance, in the southern hemisphere, the average temperature difference between summer and spring can be around 12 degrees. If we could measure CPU temps without the system running, the readings would likely match the typical room temperature at that moment. So, what was the temperature in your city seven months ago when you checked your logs?
Consider the score may reflect background processes and applications running quietly, as a fresh system typically has minimal load.
Yes, that's accurate. It wasn't the coldest winter; my room typically stayed near lower 16°C when I wasn’t using a space heater, though the CPU temperature didn’t usually rise much if I did. Regarding seasonal changes, it’s hard to rely on this as temperatures are dropping and I often open the window at night to match the room temperature with mid-winter levels while my PC runs at around 40°C. Regardless, the 5800X is a solid processor priced over $200, but in retrospect, a 5900X might have been a better choice given these conditions.