High FPS but poor performance on PC
High FPS but poor performance on PC
For two weeks now I've experienced this issue that emerged after the PC froze (it has happened multiple times), which compelled me to shut it down. When checking performance metrics with MSI Afterburner, the readings are normal (80-90 in demanding games), but they seem reduced to around 50-60, and textures load extremely slowly. I attempted using Vsync, but even though latency is very high, it doesn't fully resolve the problem as the performance remains sluggish. Could there have been any BIOS or Windows changes I'm not aware of? Or could a component have failed during the freezing?
Psdt: I've already reinstalled the operating system, and in my BIOS it's not possible to make voltage adjustments because overclocking isn't enabled. Additionally, the white light on the GPU (where the power cable is connected) hasn't turned on for months, but the PC was functioning normally until that incident.
Setup:
- Intel Xeon E5 2620 V3
- RTX 2060 6GB
- 16 GB RAM
- 1TB HDD WD Blue
- Rog Strix 650W PSU
check temperatures of CPU and GPU
check frequencies as well using hwinfo
check event viewer of windows for errors while it freezes
check windows integrity
open the command prompt as administrator and type DISM /Online /Cleanup-Image /RestoreHealth
https://www.lifewire.com/how-to-open-an-...pt-2618088
https://answers.microsoft.com/en-us...em...5b60477a93
clean boot
https://support.microsoft.com/en-us/help...in-windows
check the memory by running memtest.org usb autoinstaller (bootable USB flash drive)
check the hard drive for errors with the WDs tool and if available, update the firmware
use ddu uninstaller and reinstall the latest graphics driver
eventually replace the cmos battery
Hello, thank you for the welcome!
The board I use is a Qiyida x99-H9 with an Intel Q85 chipset. There is a Spanish page that lists all Chinese boards designed for Xeon, which I keep here:
https://placaschinas.com/
The BIOS version is 2021 American Megatrends 5.11.
Assuming you meant RAM, I’m sharing details about it:
https://www.corsair.com/es/es/p/mem...Ka...dTfHPGsXCE
During testing with CPUid, the only unusual finding was voltage fluctuations in the CPU, which also affected its parameters like frequency and usage. I observed lag when the processor experienced large high-to-low swings (ranging up to 90% and down to 20% in milliseconds). I’m not sure if this was a consistent issue before since I didn’t perform performance tests.
Also, note that the white light on the GPU remains off, which suggests it isn’t connected to the power source.
Looking forward to your reply.
The temperatures on my PC aren't perfectly accurate because the sensors on my board aren't very precise (especially the CPU, VRM, and the board itself). However, when this issue began, I changed the thermal paste, but it didn’t fix anything.
I also noticed that the CPU experiences fluctuations in voltage, frequency, and usage—rising to 90% and dropping to 30% in milliseconds.
Freezes shown in the event viewer are likely caused by insufficient power supply; a slight desk movement can cause a microcut in the power cable, though this happens infrequently. Adjusting the cable usually resolves it.
Using that command everything seems normal.
I performed a clean boot and the problems continued.
I wasn’t aware of a page issue—I’ll download it and provide an answer, but the Windows native memory test showed no RAM errors.
I plan to follow your instructions with official WD software; what I did was run a test with Crystaldisk (info and mark), and in two tests the disk read/write speeds were noticeably different—slow initially (80-75) and then improved to 135-140 for a 150MB/s HDD.
I forgot about Nvidia software—I’ll try it and reply later.
The board I use doesn’t require a battery; I always used it this way, but yesterday I put it on and there was no change.
Thanks for your help! I look forward to your responses.
I reply here since the situation is urgent. After using DDU to uninstall and reinstall the drivers, the GPU clock now shows frequency and memory fluctuations ranging from 1950 to 300 Hz, which wasn’t the case before. Before DDU, these variations didn’t occur and the lowest it dropped was 1330MHz. Did I make a mistake? Were any settings changed?
I’m not familiar with overclocking and, given my motherboard, I’m hesitant to attempt any component adjustments. That’s why I’m confused about what went wrong.
Psdt: The performance remains poor. MSI Afterburner reports 90-100 fps but doesn’t reach that speed; it consistently drops to 40 or less.
I’m waiting eagerly for your reply :c
What is being displayed could indicate a problem, as incorrect sensor readings might cause the device to slow down or shut down.
Other software indicates the PC temperatures are within normal ranges, but I rely mainly on the GPU readings since they seldom exceed 80°. This is based on several online discussions where forum users report incorrect sensor data, VRMs and CPU showing inaccurate values. On YouTube, some people tested this board using a laser thermometer: the CPU with heatsink stays below 60°, while VRMs under load reach 70-75°, and the inadequate heatsink doesn’t help. However, when using MSI Afterburner during demanding games like BF1 or WZ, I see “normal” readings (80-90 FPS) even though the actual performance is much lower (50-40 FPS). If there were issues, frame rate and timing should align with what’s visible; otherwise, the discrepancy remains.
Hi,
The answer is yes, under load the GPU rarely goes above 80°.
Sorry, according to msi ab the CPU reaches 55° at most in demanding games, but I don't trust that reading.
Right now I'm downloading 3dmark, I'll edit the comment when the tests are ready.
Edit: This is the link of the result -
http://www.3dmark.com/sn/1216594