Not from my experience. That looks like a kernel bug - boot into an older kernel and see if it persists.
Not sure what one you are on, but I had that in a couple 6.16 and 6.17 releases.
I’m on 6.16.9. It looks like it happens only on that power port. If I move the power cable to a different port it works as expected. Guess I have to do more testing and make a separate thread.
interesting - @ me when you create the new one - as well, which port causes it? I can test on mine.
I was seeing the issue on port #1 but not port #5. I’m going to try the following (probably tonight)
A) recreating the issue the same way, make sure the issue’s still present
B) Plugging directly into the mobo (i.e. no expansion card) to rule out the expansion card
C) Test the same expansion card in another port
D) Use my USB-C power meter to see what voltage the is being provided to see if it’s some sort of handshake issue.
I did try manually overwriting the TDP limit with LACT, which applied as far as nvtop can tell, but it didn’t actually change it in terms of what actually drives power usage.
1 Like
Turns out my charging ports just aren’t working and the dGPU was just responding to being on battery.
Run the battery all the way down and leave it for several minutes. Or maybe just power off your laptop for several minutes, unplugged, before starting it back up again.
Is your experience anything like what I experienced on the 3.06 beta?
1 Like
I’m seeing a lot of feedback on the instability of this release - just wondering this 3.07 release safe to install?
1 Like
No, stay on 3.05 until 4.0 is released
2 Likes
Framework says it is stable 
In all seriousness, you should probably stay on 3.05 if you have a dGPU installed.
4 Likes
I do have a dGPU. Thank you both for the answers.
I have it here, and have no issues TBH. Then again, you usually only hear those having issues.
I upgraded from 3.05 last week and I’ve been having a terrible time…finding issues. I’ve had none. Neither Windows (10, 
) nor Fedora have complained. Then again I haven’t been using it very hard, so maybe I just don’t notice? 
Now that I look back on it, the dGPU bouncing up and down in performance could’ve been an early indicator of my 1,2,4 power rail failing. I was plugged into that rail at the time, and the device didn’t charge at all between when I had that issue and when I detected the failed ports. If the ports were working intermittently, then the TDP of the dGPU would’ve been bouncing as well.
I’ll say this is a timely occurrence considering I’m currently debating getting rid of my gaming desktop and going all in on the Framework desite the performance downgrade.
I’m having a lot of issues in the new BIOS, how do I downgrade?
My dGPU is still locking to under 40W TDP at random on both Windows and Linux.
To downgrade the BIOS use this:
fwupdmgr downgrade
and follow the prompts to select which bios version to downgrade to.
1 Like
Thanks!!!
You can always upgrade/downgrade firmware from LVFS manually from GUI using Firmware application (official app from the main fwupd developer)
1 Like
You can also download the stand-alone EFI update files here:
https://downloads.frame.work/bios/Framework_Laptop_16_Amd_Ryzen7040_BIOS_3.05_EFI.zip
I wish they would just leave the older ones linked on the site until they release a bios without debilitating issues.
Thanks for the help guys, I was able to successfully revert to 3.05. Unfortunately, I’m still having issues with my FW 16 acting as if it’s running on battery when it’s not, so I guess that wasn’t it.
Think I should submit a ticket if laptop’s plugged in, below temp, and only clocking at half speed? It’ll be working normally, then some time into a game session I just lose half my frames and CPU/GPU drop in power target/clock speed.
If you are having problems with games stutter, may read this thread.
It’s a bit long, but we did get to a point where, with a custom EC firmware, written by me, users were able to play games without stutter.
If you wish to discuss it, probably start a new thread.