My self-built PC used to idle around 35°C on the CPU and 39°C on the GPU, but over the past few days both have been sitting around 40–45°C. The room remains at about 20°C, and antivirus and antimalware scans found nothing. The system has several intake and exhaust fans, plus a 360 mm liquid CPU cooler.
The bigger concern is that the display sometimes stutters and applications become extremely slow when the CPU and GPU reach roughly 50–55°C. This started around the time an NVMe SSD was replaced. The old drive randomly disconnected and sometimes prevented booting; another NVMe drive is visibly bent but still works. The drives are Corsair MP700 PRO XT models, with one installed near the graphics card.
The system uses a Ryzen 9 9900X with an overclock, an RTX 5090, and two 4 TB NVMe drives. I also updated the graphics driver recently, although the temperature increase and glitches began before that. Are the idle temperatures concerning, and could the SSD, graphics driver, or something else explain the stuttering?
4 Answers
The bent NVMe drive is a more concrete warning sign than the temperatures. Check that the motherboard standoff is in the correct position, the drive is sitting flat, and the heatsink or thermal pad is not applying excessive pressure. Monitor both drives’ temperatures and SMART/health data with a tool such as CrystalDiskInfo. Back up important data before testing further.
Also check whether the drive disappears from BIOS or Device Manager, logs storage errors, or shows rapidly increasing temperatures. A drive installed directly beside a high-power GPU may also be affected by heat or restricted airflow.
Because the problem began around a graphics-driver update and involves visual corruption, a clean driver reinstall is worth trying. Use a trusted driver version, remove the current driver with DDU from Safe Mode while temporarily offline, then install the replacement with the clean-install option. If the issue remains, the driver is less likely to be the cause.
It is also worth checking Windows Event Viewer for display-driver resets and storage errors, and testing with hardware acceleration disabled in Firefox. Since the symptoms may be caused by the SSD or another hardware fault, do not assume a driver reinstall will solve everything.
Check temperatures independently rather than judging by the moment the glitch occurs. Log CPU package temperature, GPU temperature and hotspot, NVMe temperatures, clock speeds, power usage, and drive health while reproducing the problem. A failing SSD, unstable overclock, loose connection, or power-management issue can cause lag while the CPU and GPU temperatures remain perfectly safe.
As a diagnostic step, temporarily return the Ryzen 9 9900X to stock settings and test with the questionable NVMe drive disconnected, provided the operating system is not on it. Keep backups current, especially with a visibly bent drive.
An idle temperature of 40–45°C is normal and is not dangerous for either the CPU or GPU. Even 50–55°C under light or moderate use is far below their thermal limits. The temperature increase may simply reflect a different fan curve, background activity, driver behavior, or normal sensor variation.
However, the glitches and application delays are not caused by the CPU or GPU merely reaching 50°C. The fact that they happen at the same time may mean that another component is under load or failing.
That makes sense. My previous system ran much hotter without issues, so I was probably focusing too much on the temperature number instead of the actual fault.
The affected drive is in an NVMe slot beside the GPU, and the other drive is bent as well. I’ll check the mounting and health information before moving anything, since removing the GPU and reseating the drives will take some work.