**Qubes OS version**
`R4.1`
**Are you using any [testing](https://www.qube…s-os.org/doc/testing/) repositories? If so, please list them and where they are enabled (i.e., dom0, TemplateVM, or StandaloneVM).**
Yes, installed latest updates in `current-testing` after noticing the issue on `current`.
**Are you providing feedback about a specific package or packages in testing? (If so, please list the package(s) here and [provide feedback in `updates-status`](https://www.qubes-os.org/doc/testing/#providing-feedback), linking to this issue.)**
No
**Affected component(s) or functionality**
I'm not sure, either xen or something in dom0?
**Brief summary**
I installed QubesOS 4.1 on a new laptop with Ryzen 7 4800H (without a discrete GPU) and the CPU usage is very abnormal. Even when only dom0 is running, without any other VMs and doing nothing on its own, idle CPU thermals are also very high, 60-70 degrees Celsius and laptop fans constantly spinning. I see regular CPU usage spikes in `xl top`, though not in dom0 `htop`. Starting AppVMs seems to affect this, even if they are subsequently stopped, idle CPU temperature doesn't drop to previous levels.
**How Reproducible**
For me, every time, but I haven't seen anyone describing exactly the same problems in another issue, so :man_shrugging:
**To Reproduce**
:man_shrugging:
**Expected behavior**
A smooth experience with no/low CPU usage and fan noise when idle. I tried Xubuntu on the same hardware and got no CPU usage or fan noise when idle, steady ~40C CPU temperatures after the initial OS startup.
**Actual behavior**
Laptop fans never stop spinning, going only from medium power on idle to high power when I actually do something... Performance wise, in contrast to the people in https://github.com/QubesOS/qubes-issues/issues/6055, I **don't** actually get any visual stutters or mini lock-ups after I installed the `amdgpu` package. The machine is mostly usable if it's plugged in, even if very annoying to use with the noise and probably severely under-performing. And yeah, I am not accustomed to great battery life with previous QubesOS versions, but this is more than a bit worse than usual :sweat_smile:
**Screenshots**

**Additional context**
- The problem seems to be worse after I start an AppVM, even if I stop it afterwards. If I set `qvm-prefs <app-vm> autostart false` on `sys-net`, `sys-firewall` and `sys-usb`, then CPU thermals hover around 53C. That is worse than the ~40C I got in Xubuntu, but still better than what I get after I start an AppVM. When I start another VM, even if I then immediately stop it afterwards, the CPU temperature never drops below ~62C again...
- It doesn't seem to matter if I've skipped the X login screen and logged directly into a VT. CPU thermals are still high after I start an AppVM, even if I stop it afterwards.
- `sudo htop` on an idle dom0 didn't reveal anything strange, as far as I can see - CPU usage at near 0%, while `xl top` shows occasional spikes in dom0 CPU usage... That said, the spikes are minor, climbing to a single-digit percentage every second refresh, so I'm not sure they can cause the symptoms I see :man_shrugging:
- Not sure if it's related, by I also can't suspend the laptop - trying that results in a blank screen where nothing works, not even switching to a VT with Ctrl+Alt+Fx, though the CPU fans keep spinning pretty hard all the while, and I have to do a hard shutdown of the laptop
- I have an external screen, though that doesn't seem to matter, the issue occurs even when I don't plug it in
- I also experienced some visual stutters after the initial install on the laptop, though these were fixed when I installed `xorg-x11-drv-amdgpu` in dom0
[dmesg.log](https://github.com/QubesOS/qubes-issues/files/6559520/dmesg.log)
[xl-dmesg.log](https://github.com/QubesOS/qubes-issues/files/6559521/xl-dmesg.log)
**Solutions you've tried**
- `journalctl`, `dmesg`, `xl dmesg`, `powertop`, `htop`, `xl top`, `cat Xorg.0.log`, `lspci -nnk` didn't reveal anything strange besides the minor discrepancy between `xentop` and dom0 `htop`, though I am not well versed in monitoring the performance of a xen+linux system... :disappointed:
- I experience this issue with the 5.10 kernels as well as the `latest` 5.12
- Tried adding various combinations of suggested xen and kernel boot options from https://github.com/QubesOS/qubes-issues/issues/6055 - nothing seemed to help at all, though there are so many things I don't understand there that I might have missed something important
- Tried to downgrade the dom0 xen packages to 4.14.0-4 - this caused my system to completely lock up a few seconds after I log in the GUI (though not when I did so in a VT)
**Relevant [documentation](https://www.qubes-os.org/doc/) you've consulted**
- https://www.qubes-os.org/doc/hardware-troubleshooting/
- https://www.qubes-os.org/doc/gui-troubleshooting/
**Related, [non-duplicate](https://www.qubes-os.org/doc/reporting-bugs/#new-issues-should-not-be-duplicates-of-existing-issues) issues**
- https://github.com/QubesOS/qubes-issues/issues/6055, though I am not completely sure, since the symptoms seem different
- https://github.com/QubesOS/qubes-issues/issues/6066 - I experience some of the same suspend issues, though I am not sure if they are related to this one