Reduced power limits after FW16 BIOS update to 3.07

[hariganti]

Device Information

System Model or SKU

Please select one of the following

• Framework Laptop 12 (13th Gen Intel® Core™)
• Framework Laptop 13 (11th Gen Intel® Core™)
• Framework Laptop 13 (12th Gen Intel® Core™)
• Framework Laptop 13 (13th Gen Intel® Core™)
• Framework Laptop 13 (AMD Ryzen™ 7040 Series)
• Framework Laptop 13 (AMD Ryzen™ AI 300 Series)
• Framework Laptop 13 (Intel® Core™ Ultra Series 1)
• Framework Laptop 16 (AMD Ryzen™ 7040 Series)
• Framework Laptop 16 (AMD Ryzen™ AI 300 Series)
• Framework Desktop (AMD Ryzen™ AI 300 PRO Series)

BIOS VERSION

3.07

DIY Edition information

Memory: Crucial CT32G48C40S5.M16A1
Storage: SOLIDIGM SSDPFPNU010TZ & INTEL SSDPEKNU020TZ

Port/Peripheral information

N/A

Standalone Operation

Are you running your mainboard as a standalone device. Is standalone mode enabled in the BIOS?

• Yes
• No

Describe the bug

After upgrading the BIOS to 3.07, under the following conditions, the package power limit has dropped from 60 W sustained to 40 W sustained:

• Manual battery limit of 60%
• Battery extender disabled
• Connected to 65 W USB-PD monitor
• Battery at 58-60%

Steps To Reproduce

Steps to reproduce the behavior:

1. Set system to use the performance power profile (powerprofilesctl set performance)
2. Open system monitoring application (btop)
3. Open a stress application (s-tui)
4. Start the stress test on all cores
5. Observe 40 W package power limit @ ~68 °C (20 °C ambient)

Expected behavior

Prior to this BIOS update, 60 W @ 99 °C was observed under the same conditions

Screenshots

Before BIOS update (shown during compile)

After BIOS update (shown with s-tui)

Operating System (please complete the following information):

• OS/Distribution: Arch Linux
• Version: N/A
• Linux Kernel Version: Linux cyborg 6.17.1-arch1-1 #1 SMP PREEMPT_DYNAMIC Mon, 06 Oct 2025 18:48:29 +0000 x86_64 GNU/Linux

Additional context

Previously, it appeared the battery was allowed to discharge when the performance profile was set to "performance" and the CPU was fully utilized, but it appears that battery is now maintained at nominally no charge or discharge in this scenario. The screenshot shown has a battery charge above the threshold, which is allowing discharge down to 60%, thus some power draw is shown. Ordinarily, the power observed is <0.1 W. It does seem some power is drawn from the battery, but far less than previously.

While this may have even been an intentional change for charging sources that can't fully power the laptop, I believe that the "performance" profile should allow for the laptop to use the battery, along with the external source, up to the package thermal, voltage, current, or power limits. "Balanced" and "power-saver" make more sense as targets for maintaining battery charge.

Finally, I did attempt to dig into this before creating an issue, and I believe the issue is derived from the BIOS update, but I apologize if I missed something within the OS that is actually causing this and may have changed at a coincidental time

[hariganti]

Issue confirmed to be present on Ubuntu as well (worse, in fact).

Additionally, rolling back to BIOS version 3.05 immediately resolves the issue. The CPU, under the same test conditions, quickly reaches 60 W and 4.4 GHz, averaging 20 W and 800 MHz higher than BIOS version 3.07.

[Patryk27]

Similar here - I'm observing max 30W~31W (s-tui stress and similar, e.g. compiling LLVM), connected to Sandstrøm 140W GaN charger (unfortunately lost the original fw charger 😢 -- but I did have better performance on previous BIOS version).

[xamindar]

I just want to add to this. I think this is the same issue on 4.01.
In my case I'm on bios 4.01, FW 180w psu connected.

I can reproduce with the following steps:

1. start stress (package power is already limited in my case to 35w).
2. sleep, then wake (package power now good, between 50w-65w).
3. close s-tui, wait for dGPU to enter sleep (D3cold), then re-launch s-tui.
4. See the package power is limited back down to 35w.
5. Repeat 2-4 as many times as you want to repeat it.
   Video of it here: https://www.youtube.com/shorts/rKd8w53GIxc

[Filip-K]

Hi! I don't know if it's related, but I noticed that on my machine, around the 3.07 update, the STAPM and PPT limits when on the FW 180W adapter changed from 120W to 95W for the Balanced profile. The changelog for BIOS 3.07 mentions "Fixed battery drain issues occurring with the Balanced Power Profile". However, it appears that the "fix" isn't limited to the Balanced profile, because even with the Performance profile I do not get the old 120W STAPM/PPT limits (they're 95W, even on Performance).

[hariganti]

Is BIOS 4.02 intended to address this issue? I'm not sure if item 7 (related to PROCHOT under load) was the root cause, or if the power limits from the prior comment are the root cause.

Also, @Filip-K, which OS do you use? How did you find the power limits?

[Filip-K]

Also, @Filip-K, which OS do you use? How did you find the power limits?

I'm on Bazzite 43 and I inspect the power limits with sudo ryzenadj --info.

I updated to BIOS 4.02 in the meantime and now the STAPM/PPT limits I get are 120/95/85 W on Performance/Balance/Power Saver respectively.

To get back to the core of the issue though, which is about BIOS 3.07. It could be that the lower power limits in 3.07 are a feature, not a bug, meant to prevent the battery from discharging on the Balanced power profile when using the 180 W brick. This battery drain was a very common complaint, so I can see why they would make such a change.

@hariganti Maybe BIOS 4.02 on Performance would give back the performance you lost with 3.07? I do get the old 120 W power limit with that (albeit on the 180 W brick, might not work on a 65 W PD).

[hariganti]

@Filip-K, thanks. Unfortunately, ryzen-adj doesn't work for me with lockdown mode, so I haven't been able to use that to check the power limits...

Given how tedious it was to revert the BIOS, I'm not really keen on jumping into a new BIOS that may have the same issue. Without direct confirmation from Framework that this is intended to resolve the bug where the performance profile gets limited based on power source, I'll simply wait

[Patryk27]

Given how tedious it was to revert the BIOS [...]

btw, whaddya mean? fwupdmgr downgrade is rather quick (couple of minutes?)