My PC occasionally develops severe stuttering and an FPS drop after roughly 1–2 hours in any game. The system uses a Ryzen 5 7500F, RTX 5070 12GB, 16GB DDR5-5600 RAM, and a 750W power supply. Temperatures usually stay near 65°C, system RAM usage is around 12GB, and VRAM usage is about 6GB.
When the problem occurs, MSI Afterburner shows either CPU usage or GPU usage dropping sharply. They do not always drop at the same time—sometimes the CPU load falls while the GPU remains normal, and sometimes the opposite happens. Performance eventually returns to normal. Drivers were updated about a month ago, and several troubleshooting steps have already been tried.
Could this be caused by power limits, clock behavior, motherboard temperatures, Windows, drivers, or another component throttling after extended gaming?
3 Answers
If memory usage is not reaching its limit and the normal temperatures are accurate, a background process or less obvious thermal sensor is worth investigating. Keep a hardware monitor logging clocks, power, hotspot, VRM, chipset, and drive temperatures so you can compare the readings from before and during the stutter.
Check the actual CPU and GPU power draw or TDP when the performance tanks. A sudden drop in wattage or clock speed can help distinguish a power-management limit from a software or temperature issue.
The temperatures shown for the CPU and GPU core may not tell the whole story. Monitor the motherboard VRM or MOSFET temperatures, chipset temperature, NVMe temperature, and the GPU hotspot temperature. A hotspot or VRM getting too warm can cause throttling even when the main temperature reading looks fine. Running separate CPU and GPU stress tests with OCCT could help reproduce the issue.
The motherboard is a B650E model with a built-in heatsink. I’ll monitor the VRM and other sensor readings and run OCCT tests to see whether I can reproduce the drop.