My Windows 11 PC boots normally, but every day—sometimes after 20 minutes and sometimes after about an hour—it suddenly becomes almost unusable. The mouse moves with heavy lag, windows take a long time to close, and even HWiNFO once took roughly 10 minutes to open. Task Manager shows only about 7–13% CPU usage, 45% memory usage, and 1–3% disk activity. The CPU stays near 50°C, so this does not appear to be thermal throttling.
A Windows Reset also stalled and never completed; one attempt ran for more than 24 hours. Ubuntu showed the same problem when tested, so it may not be limited to Windows.
The system uses a Ryzen 7 5700X, ASUS PRIME B550M-K motherboard, 16 GB of DDR4 memory, an MSI RTX 3060, and a Kingston NV2 500 GB NVMe drive. The BIOS is updated. CrystalDiskInfo reports the NVMe drive as healthy with no media errors and a full PCIe 4.0 x4 link. Memory tests passed, DISM and SFC found no issues, and I have reseated the RAM, graphics card, and M.2 drive. I also completely disassembled and rebuilt the PC apart from the CPU and cooler.
The monitor briefly lost signal several times, and replacing the HDMI cable stopped the flickering. The computer then worked normally for a few days before the severe slowdowns returned. The external HDD used for backups has been disconnected, but that made no difference.
What would you investigate next? I am especially wondering whether the motherboard, GPU, or power supply could cause this kind of system-wide lag without showing high usage or temperatures.
3 Answers
Since the problem occurs outside Windows, stop spending too much time on SFC, DISM, or Windows Reset. Those tools cannot rule out a failing motherboard, GPU, PSU, or NVMe controller. Disconnect all extra drives and peripherals, boot a live Linux environment with the NVMe removed if possible, and see whether the lag still develops.
If it does, replace parts in a practical order: borrow a graphics card first, then test a known-good PSU, and finally consider the motherboard. If the issue only happens with the NVMe installed, replace that drive even if its health percentage and SMART data look good—those checks do not guarantee that a drive or its controller behaves correctly under real workloads.
The graphics card and motherboard are high on my list, especially because of the earlier signal drops, although the HDMI cable may have been a coincidence or simply made the connection more stable. Try a different monitor or TV, another display output on the GPU, and a spare GPU if you can borrow one. Also test with only the essential keyboard and mouse connected.
If swapping the GPU and display setup does not change anything, the motherboard or PSU becomes more suspicious. A PSU can produce instability without immediately causing shutdowns or obvious voltage warnings, and a motherboard problem may not appear in normal temperature or usage readings. A known-good PSU swap is more informative than software voltage readings.
This sounds like a hardware fault or a low-level I/O problem rather than ordinary CPU, memory, or temperature overload. Since it happens every day but at inconsistent times, keep an eye on Event Viewer and Reliability Monitor when it starts. Look for disk, display-driver, WHEA, USB, or Kernel-Power errors, and check whether a scheduled task or background process appears at the same time.
The most useful isolation test would be to boot a Linux live USB and run from that without using the NVMe drive. Since Ubuntu also showed the problem, Windows and its drivers become much less likely. If the issue still occurs in that environment, test with another GPU, disconnect every nonessential USB device, and try a known-good power supply. A PSU or motherboard fault cannot really be proven by software monitoring alone; swapping in a known-good part is the practical test.
If the problem disappears with the NVMe removed or under a clean Windows installation, replace the drive or reinstall Windows from USB rather than relying on Windows Reset. Make sure important data is backed up before changing storage or BIOS settings.
The slowdown happens every day, usually 20 minutes to an hour after boot. I disconnected the external HDD and tested Ubuntu, but the issue still appeared. I will try another GPU, check the logs, and perform a clean installation from USB.
The display issue could point toward the GPU or its port, but it would not normally explain the entire desktop becoming unusably slow by itself. I would treat it as a clue and still test the PSU and motherboard after trying another GPU.