Mouse cursor laggy with certain high-CPU workloads?

Discussion in 'Processors and motherboards AMD' started by Espionage724, Dec 3, 2021.

  1. Espionage724

    Espionage724 Master Guru

    Messages:
    312
    Likes Received:
    152
    GPU:
    RTX 3060 XC BLACK
    I have an ASUS X470-PRIME PRO motherboard and CORSAIR RGB HARPOON USB mouse. I've noticed at times (rarely randomly, but often with certain CPU stress tests) the cursor will get visibly laggy. It's hard to reproduce this outside of the following test, as it doesn't seem like prime95, CPU-Z's, nor OCCT's stress tests cause this, and it doesn't entirely seem like just a 100% CPU usage issue.

    I've owned Skylake, Kaby Lake, and Coffee Lake computers with this same mouse, and don't recall ever seeing my cursor ever be an issue. I know Ryzen is known for USB issues (still run into them today with VR), which has me thinking this is a Ryzen-specific issue, but interestingly even running the mouse through a different USB controller on a PCI-E card still can reproduce this stutter.

    This seemingly affects all of winsat's cpu tests (encryption, encryption2, compression, compression2).

    The easiest way I can reproduce this is on Windows by using the CPU encryption winsat test (command prompt):

    Code:
    winsat cpu -encryption
    1. Can anyone else reproduce cursor stutter/lag while that test is running?
    2. Does anyone have any idea if this can be fixed?
     
    Last edited: Dec 3, 2021
  2. mbk1969

    mbk1969 Ancient Guru

    Messages:
    11,683
    Likes Received:
    9,678
    GPU:
    GF RTX 2070 Super
    Seems natural to have mouse lags with high CPU load.
     
  3. user1

    user1 Ancient Guru

    Messages:
    1,855
    Likes Received:
    685
    GPU:
    hd 6870
    from what I remember this is a potential consequence of having either using a software cursor , a poor implementation of a hardware cursor or bad scheduling(my money is on this).It shouldn't happen if you're not running the workload with realtime priority. I would say to report a bug, but the chances of anyone remotely capable of fixing it, actually seeing it, is very low
     
  4. aufkrawall2

    aufkrawall2 Maha Guru

    Messages:
    1,244
    Likes Received:
    281
    GPU:
    3060 TUF
    Yes, it's Windows scheduling. When you run some intense number crunching on all available threads with normal process priority, Windows tends to semi-freeze much more than Linux.
    I experience this all the time on my garbage Gemini Lake 2C 2T notebook where I got dualboot Windows and Linux.
     

  5. Espionage724

    Espionage724 Master Guru

    Messages:
    312
    Likes Received:
    152
    GPU:
    RTX 3060 XC BLACK
    That's what I would think, but my cursor is fine when running a general stress-test (prime95, CPU-Z, OCCT).
     
  6. Ener

    Ener New Member

    Messages:
    4
    Likes Received:
    2
    GPU:
    1660S
    I can confirm circling my cursor while winsat cpu -encryption is running results in awful stuttering. System is Ryzen 3800x with asus x370 crosshair which means it was left for dead well before the USB AGESA fix. Your motherboard should have received the fix though.
     
  7. mbk1969

    mbk1969 Ancient Guru

    Messages:
    11,683
    Likes Received:
    9,678
    GPU:
    GF RTX 2070 Super
    All tests can load different amount of cores. I mean core 0 services interrupts usually, and if it is not loaded by the test then there will be no stutters.
     
    aufkrawall2 likes this.
  8. BuildeR2

    BuildeR2 Ancient Guru

    Messages:
    2,926
    Likes Received:
    177
    GPU:
    MSI 2080 Ti GX Trio
    1. Yes, it happens to my system
    2. No, but I've tried several things with no luck so far

    My story is much the same as yours, I've owned systems built on Sandy Bridge, Ivy Bridge, Haswell and Skylake and never had the mouse do this sort of thing. With my MSI X570 MEG ACE/Ryzen 7 5800X gaming system and MSI B550 Mortar/Ryzen 5 3600X audio computer I get the mouse thing. I've tried all the rear and front USB ports but they all suffer. On a slightly interesting note, I've noticed that when launching the bigger AAA games from Game Pass via the MS Store or Xbox app the system will get this massive mouse stutter situation. Seems like @mbk1969 is onto something about how different things can load cores differently. Also @Ener I have the latest BIOS for both motherboards and still have this issue. I was hoping they fixed it but no luck.
     
  9. aufkrawall2

    aufkrawall2 Maha Guru

    Messages:
    1,244
    Likes Received:
    281
    GPU:
    3060 TUF
    Try Linux and see the hardware cursor stutter go away in most cases. Well, at least with AMD GPU (Nvidia Linux driver is a no go) and corresponding graphical environment (e.g. Xorg with xf86-video-amdgpu).
     
  10. Espionage724

    Espionage724 Master Guru

    Messages:
    312
    Likes Received:
    152
    GPU:
    RTX 3060 XC BLACK
    I notice it more often with GNOME on Wayland when just sitting on a blank desktop, but I have a feeling it's being excalibrated by the cursor not being on its own thread like it is on Xorg. It's fine with GNOME on X, but I'm not sure how to reproduce the CPU load on Linux that causes it on Windows (prime95 on Linux doesn't cause the issue).
     

  11. aufkrawall2

    aufkrawall2 Maha Guru

    Messages:
    1,244
    Likes Received:
    281
    GPU:
    3060 TUF
    Yes, Gnome Wayland still has issues with shell/compositor stalling cursor updates. It's pretty bad, I've lost hope for it. Hopefully Plasma Wayland won't have this issue when they switch to atomic cursor API.
     
  12. Guru3Dmember

    Guru3Dmember Active Member

    Messages:
    52
    Likes Received:
    4
    GPU:
    6Gb
    Is your mouse on the CPU usb I/O, or the south bridge USB hub?
     
  13. BuildeR2

    BuildeR2 Ancient Guru

    Messages:
    2,926
    Likes Received:
    177
    GPU:
    MSI 2080 Ti GX Trio
    This is a good shout, as I was wondering the same thing when I started having this issue. I've tried every single USB port and have the same end result. No change before or after the firmware that was touted to solve other USB issues. :(
     

Share This Page