This is so tiresome. My initial positive view on Framework is seriously damaged by this BIOS drama, which is since day 1 I got the device.
Can confirm. 3.07 still stuttering with Delta 240W on High Performance. No problems on “Balanced” for now.
1 Like
After comparing the changes between versions 3.05, 3.06, and 3.07, we found no modifications related to thermal control. The PROHOT signal originates from the CPU when it detects high temperatures. A significant change in versions 3.06 and 3.07 was the update to the AMD PI code to meet security requirements, but other products with this same update are not experiencing a similar issue. We are running the heavy loading to see if we can meet the problem.
2 Likes
To clarify the PROCHOT_CPU and PROCHOT_GPU flags being set, they trigger for me when the FW 180w PSU is connected. This happens on 3.05 as well. I think the new issue for 3.07 is the EDC_CPU being triggered as well. If you say there were no modifications then that makes sense as all three bios versions have this problem.
Is it possible mine has been defective this whole time? Are others not getting these triggered with the 180w PSU connected?
1 Like
On 3.06 with both FW 180W and Delta 240W, if I disconnect the PSU the laptop goes in “power saving” mode, and there isn’t a way to take it out of it unless I reboot.
CPU-Z (plugged in), stress: ~6000pts
unplug
~4200pts
replug
still ~4200pts, clock ~2800MHz.
1 Like
We performed a CPU and GPU stress test using the configuration and steps below. No PROCHOT was triggered. Please let us know if you identify any steps or configuration details that might be causing the issue.
Test Environment
- Target machine: Framework Laptop 16 7940HS + AMD 7700S
- BIOS version: 3.07
- Driver version(s): Download Driver Bundle (v2.06) 2025-09-08
- OS: Windows 24H2
- Power adapter: 180W
Test Tools
- FurMark (latest version) – for CPU Burn abd GPU stress test
- HWInfo64 / TaskManger (Super + X then T)
- CPU package temperature
- GPU temperature
- Power draw (CPU + GPU + Adapter)
- Thermal throttling indicators
Test Steps
- Boot system with the test BIOS and drivers.
- Ensure system is at room temperature (~28°C), fully charged to almost 100% and idle for 5 minutes
- Set Power Mode as Performance mode (Both On battery and Plugged in)
- Unplug 180W adapter
- Launch monitoring tool(s) - Task Manager and monitor for temperature, power, and throttling events.
- Start FurMark GPU and CPU stress test at:
- GUP: Default RUN as resolution: 1920×1080
- CPU: Default thread (16 or 24)
- Duration: at least 20 minutes or until throttling occurs.
- Observe below parameters:
- CPU/GPU temperatures
- CPU/GPU power consumption
- Whether PROCHOT flag is triggered
- Any thermal throttling events (CPU/GPU clock drops, power capping).
- Stop test after 20mins.
- Repeat Steps 5~7 with AC plugged and monitor the battery charging ststus
| Power Mode | BIOS Version | Driver Version | Test Duration | PROCHOT Triggered | Notes |
|---|---|---|---|---|---|
| On Battery | 3.07 | Driver Bundle (v2.06) 2025-09-08 | 20 min | No | |
| Plugged in | 3.07 | Driver Bundle (v2.06) 2025-09-08 | 20 min | No |
This is with 3.07 bios that’s “hidden” on the downloads page?
it is not available on download page, but link was shared by one of the user who received it from support A call on 240w adapter - #474 by Guest103
1 Like
Furmark is a bit of an outliar in my personal experience… I haven’t tester 3.07 extensively yet, but I was able to pull full 100W on the dGPU in Furmark and have 3D Mark Steel Nomad stutter on the same testing config when testing the 240W Delta brick.
TLDR (English isn’t my first language) furmark might be able to be more “stable” (power consumption wise) than other workloads/benchmarks/real world games.
t’s still in alpha because we need to confirm this issue. While we don’t usually provide alpha versions to users, this issue was reported on GitHub, so we shared a testing link in github.
I just tried executing the same exact steps, but in “balanced” mode. BIOS is 3.07, drivers 2.06 (all latest).Once the 180W brick is unplugged, the laptop is locked at 2.5GHz even when re-plugged in
CPU-Z score drops to ~3700, and CPU maxed out at <5W.
Looks like the only fix is reboot. Happens every time I unplug the cord.
EDIT: tried the same steps in “high performance”, so exactly the steps you described.Same behaviour: looks like the CPU is capped at ~5W and the GPU at 40W, no matter how I change the power settings, unplug the cord etc.
EDIT2: got tired and reinstalled Windows 11 from scratch.
re-installed BIOS 3.07 after downgrading from 3.06, and installed 2.06 drivers (latest).
Repeating the test, no performance cap with 180W after multiple unplugging events.
With the Delta 240W the problem persists tho. But the situation with the stock 180W charger looks promising for now! 
2 Likes
The 240W test is being run with a new BIOS that supports the Framework Laptop 16 with either the AMD 7700 or Nvidia 5070. We will do the 3.07 beta released first.
6 Likes
Thanks, amazing! Can’t wait to get my hands on the new GPU! So happy to see Framework fixing the firmware, one step at a time! 
If i’m not wrong, you’re using `amdgpu_top` or similar to detect PROCHOT_CPU and PROCHOT_GPU. Are you sure that this software reports the throttle flags correctly?
I have no reason to believe it to be incorrect. If you have evidence this tool is, or can be, incorrect, I would be interested in that info.
Of note, mangohud also lists cpu temp throttling while playing games, even when the cpu and gpu temps are both in the 70s.
So I am seeing two places reporting throttling.
I can confirm that LACT has a known issue that reports “hotspot” on the dGPU no matter what - and is an issue in their github with certain versions of AMD cards.
Not sure how that relates to AMD_GPU top or the others.
1 Like
I have at least seen those flags a lot when there was no obvious throttling going on on the 13.
1 Like
hello im on windows and i don’t know how to make the dGpu use the highest clocks everytime i play a game it will crash and make my laptop restart and i think the issue is from the dGpu not having max power required (im using the 180w adapter) if you can help me please
I am not a Windows guy by far.. I haven’t used one regularly in over 10 years. So… I am probably wrong, and you should research this before making any changes:
Messing with clocks, wattage, and gpu configuration is not recommended if you do not know what you are doing. You can potentially break your laptop if you do the wrong things.
While unlikely with the above tool - you should still educate yourself on the changes you are potentially making - no one is responsible for you turning your $2500 laptop into a paperweight except yourself