The cursor and apparently the system itself freeze even when idle on new hardware

[underik15a202ea]

So, I've been tearing my hair out for over a month (I just love it) after upgrading three components on my PC (CPU + motherboard + RAM). Detailed specs of my Vista beast:
◘ Processor - AMD Ryzen 9 9950X3D, ~5.7GHz (previously an Intel i9 12900KF);
◘ Cooling - Corsair iCUE H170i Elite Capellix, 420mm, 3x140mm;
◘ Graphics Card - ColorFul iGame GeForce RTX 3080 Ti 12GB Vulcan OC-V;
◘ RAM - Adata XPG Lancer RGB DDR5-6400Mhz (XMP 32-39-39-89), 2x32GB (previously G.Skill Trident Z5 RGB DDR5-5600Mhz (XMP 36-36-36-76), 2x16GB);
◘ Motherboard - MSI MPG X870E Carbon WIFI (previously Gigabyte Z690 Aorus Master);
◘ SSD - Samsung 980 PRO NVMe M.2, Samsung 970 EVO Plus NVMe M.2, Samsung 870 EVO 2.5" SATA;
◘ HDD - WD Blue 4TB
◘ Power Supply - Deepcool PX1200G, 1200W;
◘ Sound Card - Creative Sound BlasterX G6;
◘ OS - Windows 11 (25H2). The same issues occurred on 24H2.

Problem Description: Occasionally and rarely, strange and incomprehensible mouse cursor freezes and teleports occur, even on the desktop (most likely affecting not just the cursor, but the entire system). However, this doesn't happen often enough for me to record it on video. However, I can definitely see these freezes when I click "Power" in Windows settings (I've uploaded a video showing the issue to Yandex.Disk - https://disk.yandex.ru/i/iTqBbOsQcip84Q). I don't remember, but I'm also not sure how it worked on the previous processor. A friend of mine has Windows 11 and doesn't have this problem on his Intel processor, I confirmed that. Otherwise, I haven't seen any issues, meaning I haven't noticed any obvious issues in games or anything else under load.

Immediately after installing the new components, I noticed this issue, along with several others, which were resolved by a clean reinstallation of Windows 11 (canonically, via the BIOS and formatting the system drive). The freezes were also resolved. Freezing can occur on the desktop, as well as when switching or dragging tabs or windows, or when browsing the browser, and very rarely in games. The freezes last literally a few milliseconds, but they are noticeable, especially when trying to open a file or folder and accidentally teleporting the cursor.

I tested the PC in various settings. Benchmarks and stress tests, tested the memory in Testmem5 with different configurations. Everything is fine.

What I tried:

- Updated and reset the BIOS (many times, same thing with a bare BIOS). I updated the BIOS immediately after installing new components: first, it was an A64, then and now, an A70 (I also tried an A27);

- Tried a different mouse, but the problem clearly wasn't with it.

- I even tried replacing it with old RAM to rule out defects, same thing. I disabled all USB ports except the mouse and keyboard;

- Disabled unnecessary Bluetooth/Wi-Fi, and the first Ethernet controller in the BIOS (there are two on this motherboard, 2.5GBe and 5GBe);

- Tried all power types in the Power Options;

- Installed a swap partition with auto-sizing on a different SSD;

- Hopeless, but to rule it out, I ran the SFC /scannow command;

- Naturally, I updated and did a clean install. Drivers for the video card and chipset;

- I monitored the event log and didn't see anything suspicious that could have caused this, only trivial and typical unimportant errors and warnings that everyone has;

I've been surfing a lot on both Russian and foreign websites—it seems like someone has had similar problems, but they've solved them using methods that didn't work for me. Otherwise, either there's no information, or people describe it differently, and the search engine doesn't consider their search query for me. Sorry for the long list, but I wanted to describe the problem in detail. Unfortunately, I'm not a rich daddy and can't afford to replace them with new ones just to test and rule out problems with the processor and motherboard.

I'm sure the solution is somewhere in the background (as is usually the case). I hope to see constructive comments here with real suggestions, not just trivial advice. I'm even willing to offer a cash reward for a solution that truly helps me. The problem is complex and far from obvious.

UPD. The problem was found, and it was connected to two monitors connected to the graphics card (DisplayPort or HDMI—it doesn't matter). When connecting one monitor, there were no stutters. The only question was why, as the previous processor didn't have such issues...

 

[citay]

I'm going at it from all angles, trying to find other users who have this problem in common with you, in an effort to find a solution. The more i know about all the pieces of the puzzle, the higher the chance of solving it.

I know you didn't have this problem with the last system, using this card and these monitors. So we know for sure the new board/CPU have to be involved somehow. However, your graphics card and your monitors might be triggering the problem in the first place somehow. To research this, i have to know the monitor models too, and how they're connected.

In case they both have DP and HDMI input, have you tried swapping the cables around?

 

[underik15a202ea]

I'm going at it from all angles, trying to find other users who have this problem in common with you, in an effort to find a solution. The more i know about all the pieces of the puzzle, the higher the chance of solving it.

I know you didn't have this problem with the last system, using this card and these monitors. So we know for sure the new board/CPU have to be involved somehow. However, your graphics card and your monitors might be triggering the problem in the first place somehow. To research this, i have to know the monitor models too, and how they're connected.

In case they both have DP and HDMI input, have you tried swapping the cables around?

I understand, but I still blame the motherboard. Yes, I tried DisplayPort + DisplayPort, DisplayPort + HDMI

 

[citay]

Well, the motherboard seems to have some part of it, but we don't know how exactly. So for now i prefer to troubleshoot from the end where we do know the problems are triggered from: Two monitors. You have two different unknown monitors. If you can borrow a third monitor, you can test it in combination with each of your existing monitors individually. This way, if the problem happens with each of the two combination (borrowed one + monitor 1, vs. borrowed one + monitor 2), your monitors would be ruled out, because it's proven that it would happen with any two monitors used at the same time. But if it only happens in one of the two combinations, then we would have an interesting new lead.

 

[underik15a202ea]

Well, the motherboard seems to have some part of it, but we don't know how exactly. So for now i prefer to troubleshoot from the end where we do know the problems are triggered from: Two monitors. You have two different unknown monitors. If you can borrow a third monitor, you can test it in combination with each of your existing monitors individually. This way, if the problem happens with each of the two combination (borrowed one + monitor 1, vs. borrowed one + monitor 2), your monitors would be ruled out, because it's proven that it would happen with any two monitors used at the same time. But if it only happens in one of the two combinations, then we would have an interesting new lead.

Unfortunately, there is no way to find a third monitor) I understand that I need to proceed by elimination. Ideally, I'd buy another motherboard and test it on that, but if the problem persists, I'll just be wasting time and money (even if I sell the current one).

 

[citay]

Yes, replacing the board has the risk of not solving it. You can open a ticket with MSI if you haven't done so yet, but often times if they're unsure about the cause too, all they can do in the end is an RMA. And if this is some weird issue that affects this platform sometimes, then replacing the board would be futile. The cause has to be located more precisely. But you can check what MSI reply.

BTW, i cannot watch your video, the website throws an error, "SSL_ERROR_RX_RECORD_TOO_LONG", so I've only been going by how you described the problem.

 

[underik15a202ea]

Yes, replacing the board has the risk of not solving it. You can open a ticket with MSI if you haven't done so yet, but often times if they're unsure about the cause too, all they can do in the end is an RMA. And if this is some weird issue that affects this platform sometimes, then replacing the board would be futile. The cause has to be located more precisely. But you can check what MSI reply.

BTW, i cannot watch your video, the website throws an error, "SSL_ERROR_RX_RECORD_TOO_LONG", so I've only been going by how you described the problem.

I'll try filing a ticket. Thanks and sorry for wasting ur time. Where can I upload the video?

 

[citay]

Youtube perhaps?

In the meantime i had another idea. What if you use one port from the iGPU, and the other from the graphics card? Maybe that could be a viable workaround.

 

[underik15a202ea]

Youtube perhaps?

In the meantime i had another idea. What if you use one port from the iGPU, and the other from the graphics card? Maybe that could be a viable workaround.

Your decision with the iGPU will be more problematic and conflict-ridden, I assure you. I don't plan to use the integrated graphics card, so I immediately disabled it in the BIOS.

 

[underik15a202ea]

Your decision with the iGPU will be more problematic and conflict-ridden, I assure you. I don't plan to use the integrated graphics card, so I immediately disabled it in the BIOS.

In short, the overall system stuttering issue and the specific stuttering issue after pressing the "Power" button in Windows Settings are two different things. The second issue disappeared with one monitor, but the first one remains. And it's the same frequency as it was with two monitors.

 

[citay]

Your decision with the iGPU will be more problematic and conflict-ridden, I assure you. I don't plan to use the integrated graphics card, so I immediately disabled it in the BIOS.

My friend uses it for years now in his basement which we turned into a music recording studio, but where he also games. One monitor on the graphics card is for gaming, the other one on the iGPU shows Youtube videos or whatever else. This is actually a good solution, because when the GPU is under full gaming load, it doesn't also have to accelerate video playback on a second screen. There are no problems, apart from two drivers needing updates (his is an Intel + NVIDIA combo).

In short, the overall system stuttering issue and the specific stuttering issue after pressing the "Power" button in Windows Settings are two different things. The second issue disappeared with one monitor, but the first one remains. And it's the same frequency as it was with two monitors.

Check with LatencyMon if it shows high latencies: https://www.resplendence.com/latencymon

 

[underik15a202ea]

My friend uses it for years now in his basement which we turned into a music recording studio, but where he also games. One monitor on the graphics card is for gaming, the other one on the iGPU shows Youtube videos or whatever else. This is actually a good solution, because when the GPU is under full gaming load, it doesn't also have to accelerate video playback on a second screen. There are no problems, apart from two drivers needing updates (his is an Intel + NVIDIA combo).



Check with LatencyMon if it shows high latencies: https://www.resplendence.com/latencymon

Okay

 

[underik15a202ea]

My friend uses it for years now in his basement which we turned into a music recording studio, but where he also games. One monitor on the graphics card is for gaming, the other one on the iGPU shows Youtube videos or whatever else. This is actually a good solution, because when the GPU is under full gaming load, it doesn't also have to accelerate video playback on a second screen. There are no problems, apart from two drivers needing updates (his is an Intel + NVIDIA combo).



Check with LatencyMon if it shows high latencies: https://www.resplendence.com/latencymon

The 10 minutes passed without any problems. Before the test, I turned off all background apps and did nothing for the entire 10 minutes. I think that's how it's supposed to be.

 

[citay]

Yes, this is a very good result. So if there are still weird "jumps" of the mouse cursor with such a result, it's not something that Windows is really aware of.

 

[underik15a202ea]

Yes, this is a very good result. So if there are still weird "jumps" of the mouse cursor with such a result, it's not something that Windows is really aware of.

So it turns out the problem is hardware?

 

[citay]

It's something so low-level that Windows apparently doesn't really register it. Possibly something with the BIOS, or even a BIOS setting. My friend with the recording studio, we updated his BIOS a couple months ago, and suddenly he had dropouts in his audio. After much trial and error, i discovered that it had to do with the settings "Native PCIe Enable" (can be called "PCIe Native Power Management" on newer boards) and "Native ASPM", they both defaulted to Disabled in that BIOS version, and after i enabled them, the problem was gone.

It could be something as simple as this sometimes. Although this was on Intel, and i don't know if this is available on AMD-based boards, plus it's just an example.

But if it's a BIOS bug, it won't be this simple. Then MSI have to try to reproduce it in their labs, and work on a fix.

 

[underik15a202ea]

It's something so low-level that Windows apparently doesn't really register it. Possibly something with the BIOS, or even a BIOS setting. My friend with the recording studio, we updated his BIOS a couple months ago, and suddenly he had dropouts in his audio. After much trial and error, i discovered that it had to do with the settings "Native PCIe Enable" (can be called "PCIe Native Power Management" on newer boards) and "Native ASPM", they both defaulted to Disabled in that BIOS version, and after i enabled them, the problem was gone.

View attachment 207239 View attachment 207240

It could be something as simple as this sometimes. Although this was on Intel, and i don't know if this is available on AMD-based boards, plus it's just an example.

But if it's a BIOS bug, it won't be this simple. Then MSI have to try to reproduce it in their labs, and work on a fix.

Thanks, I'll try that. At least my sound isn't dropping out.