You can go lower on the lefthand side.40% ist still noticable loud for “nonuse” temps. As i wrote keep it below 40% up until 60C.
You want the hysteresis ar 4s as any little temp jump shouldnt trigger a big fan change.
Also, does the app have to have an open window for the curve to be running in the background?
there is no need to keep browser open. it is used to configure background process which is monitoring temperature and changing fan speed
The default behavior of the fans set by the Framework team (auto mode) doesn’t go all the way up to 100% so I’d say it’s not strictly necessary to go to 100% on 100C. It gets super loud at 100%.
With that said, it’s definitely safer to have fans running faster than slower.
The laptop is designed to hit 100°C (from my testing at least), but I’d recommend NOT keeping it there for long periods of time. So until I release the TDP and Thermal Limit adjustments (WIP), you’re kinda stuck with higher fan speeds.
If you go lower than 10% the fans completely stop, so I’d suggest staying above that to not have a bunch of stop-and-starts, which may reduce longevity (don’t quote me on that; maybe it’s fine to stop/start).
Also, if you’re curious about how I have my cuve set up, here’s a screenshot.
Keep in mind, I have access to the thermal limit stuff early because I’m the dev
. So my laptop never crosses 90°C, you might want to set a more aggressive slope at the higher temps.
To summarize… Everything is OK as long as you’re keeping an eye out on your live temps and other metrics like TDP, which you can do with software like HWINFO.
I have one question. is it normal that framework-control-service.exe is causing 4-5% constant CPU utilization when switching to “curve”. Here are my settings
If you’re using the FW 16 that’s definitely not normal. My service uses between 0% and 0.1% at all times…
Could you send me the contents of the “C:\Program Files\FrameworkControl\FrameworkControlService.out.log” and the “FrameworkControlService.err.log”?
Thank yo for help. uninstalling and installing one more time solves the issue
Appreciate you making this web app and also making it his intuitive. Now I can keep the fan speed higher at low temperatures to not burn my lap, and also I always thought that the stock profile spun fans at its maximum at maximum load. The 100% fan speed blows a lot of air!
Thank you for the app! I encountered a problem: When I set the control to “curve” and use the 7700S graphics, the temperature of the graphics card does not influence the fan. Therefore the graphics card can get pretty hot. When I switch to automatic the fan ramps up immediately. I have set a high cooling curve, where the fan is at about 50% at 57°C and 100% at 80°C. Am I doing something wrong? If not: may I suggest to implement that the fan speed is never allowed to drop lower than the automatic value? This would be a safety measure to not run the system to close to the edge.
Right below the curve, it says “CPU” or “APU”. That’s the sensor used to enforce the curve. The program right now does not support dGPU temperature monitoring, so it will noto care about the 7700S’ temperature.
The stock curve (Framework’s) uses a combination of CPU and dGPU sensors, so it ramps up if you’re using either component.
I had one other person tell me about this. Will get the option to choose the dGPU as well as the maximum of all three.
I may or may not wait for the next major update for this but keep an eye out for updates!
Thanks for clarification.
Thank you so much for all your effort!
Finally got this installed last night! Didn’t have time to mess with anything but I’m excited to try it out! I really think having the option of the dGPU being an option is certainly a good idea. Especially for those of us who have and use it.
This works for my FW13 AI 300 on 3.04 too. I fucking love you. No more teppanyaki human thighs when I actually use it on my lap ![]()
I’m glad that it’s working on the AI 300, thank you for giving it a shot! ![]()
Another vote for adding the dGPU temp setting if possible, and using the maximum of all three (CPU, iGPU/APU, and dGPU) when on curve.
Out of curiosity, do you have any timeline on any future updates? You had a flurry of updates at the end of August and then the last one at the beginning of Sept. Do you plan to accept PRs if offered (I wouldn’t mind sharing some of the work)?
What you’ve provided already is wonderfully helpful. Thank you!
I’m working on that, been pretty busy with other life stuff, two weeks ago got all four of my wisdom teeth removed
, and today I needed to change the brakes on my car ![]()
To answer your questions:
- No, I don’t have a strict timeline and can only work on this project in my spare time.
- I might merge the simple and straightforward PRs, but I am super picky with the code I commit and code I accept from others, so no promises for stuff to be merged.
If you want to help out, feel free to review and post your suggestions to this upcoming PR. It’s a big release that adds TDP and Thermal Limit adjustments and other minor stuff.






