Mmmmm not so much. The reality is somewhere in between that and it’s not an issue.
This is when the fans of maxed with severe work load…that is not normal.
This is most likely non optimized fan curves. Fans do not just go from 0 to 100%. The curve is just as important as the noise level along with what frequency they are at. Frequency is the difference between normal background noise and wanting to throw something across the room.
I saw the exact same issue with my Surface Pro 4 when it came out. Within 2 months the firmware was updated and the drone went away with more proper turned fan curves.
The current dGPU module doesn’t include it, but it is possible for future modules. The Expansion Bay connector has pins to support sending power from the Expansion Bay to the mainboard.
I posted this in a related thread on Reddit, but for more detail:
We’re certainly committed to making sure Framework Laptop 16 is a solid product. For the total list of issues that we’ve resolved on customer units since the press units we manufactured in December:
High frequency noise from Mainboard - We identified an incorrect capacitor value that results in a high pitched noise during high load while using a 28V or 36V power adapter. This is resolved on customer units.
Buzzing or chirping noise from Graphics Module - We identified a scenario where the inductors on the Graphics Module can buzz under high variations in load. We’ve updated the Graphics Module inductor assembly on customer units to resolve this.
CPU thermal module performance - Our thermal module supplier improved their vapor chamber soldering process, which reduced thermal resistance. While this was only intended to improve manufacturing yield, it actually ended up improving thermal performance too. All press units passed the same pass/fail criteria that we use for CPU performance on customer units though, so we consider press unit CPU benchmarking to be a fair representation of what customers will receive.
Liquid metal barrier adjustments - We made some adjustments to the liquid metal application process to prevent any leakage risk on customer units.
Cold GPU performance - For GPU benchmarks, on a cold first run the scores may be lower than subsequent warm runs. We resolved this through a BIOS update that we provided to reviewers partway through the review cycle, which some reviewers may have benchmarked on older BIOS. We know that LTT was on the newer BIOS in their benchmarking.
DPC_Watchdog_Violation blue screen - There was a system stability issue that occurred primarily when scrolling the touchpad that could result in a blue screen. This was the issue that The Verge ran into, and we’ve since resolved it in the BIOS that is on customer systems.
Speaker weighted to left or right channel - There is a bug in the smart amp DSP in which the left or right channel may be attenuated at certain times. We saw a reviewer specifically call out that audio sounded shifted. We’ve found the root cause of this, and we’ve resolved it in the driver on customer systems.
Touchpad Module sliding friction - The mechanical structure that the Touchpad Module slides into is slightly deformed on some press units, resulting in higher sliding friction. This is resolved on customer systems.
Display alignment - On some press units, the display was slightly misaligned in a way that resulted in the bezel covering the edge of the active area. We bypassed screening for this during press unit manufacturing, but are checking this during production of all customer units.
Minor fit and finish in the Input Modules - There are some mechanical alignment refinements in customer systems that improve the alignment of the pins in the Mid Plate to the alignment holes in the Input Modules, as well as reduce visible gaps along the top edges of Spacer Modules. There are also minor improvements in the flatness of the Touchpad Module and Touchpad Spacers in customer systems.
Remaining open questions:
Keyboard deflection - This is an unusual one, because we’ve seen pretty consistently positive feedback on the input deck feel from most of the reviewers, but also specifically saw LTT’s video and the workaround they applied. We’re investigating whether there could have been either an issue on that unit, or a scenario that can result in the mid plate not being flat. In either scenario, if we find that there is an improvement we can apply on this, we will do so, and ship out any parts necessary for that to customers.
I got faith in you guys. Even if something is found after systems ship I know you folks will fix it. Framework is the only company I’ve run across that keeps its customers informed about the good and the bad. That honesty is why a lot of us are going to give you the benefit of the doubt.
We kept the newsletter itself pretty high level since it goes to an extremely broad audience, but we posted the specific differences elsewhere in this thread. We’ll also share this detailed list out to Framework Laptop 16 customers in the next shipping update email we send.
There was indeed a specific firmware issue that resulted in a blue screen on Windows in some scenarios, which we’ve since fixed. It seems most of the reviewers didn’t run into it, but The Verge did. Actually, in our internal testing, I was the only one in our team who was seeing the same blue screen, which helped us debug and root cause it.
It’s been great to see the positive feedback from Phoronix on Linux though.
On the vapor chamber issue that we resolved on customer units, the problem was unit to unit variation. Both LTT and The Verge received units from the same batch, but it is possible that their samples behaved differently. The Verge’s heatsink did pass our test criteria though in the factory, so we of course accept their results and feedback as is.
We are going to continue to tune fan curves. We had one fan curve update during the press review cycle, which means some reviewers were on a newer curve and some may have been on an older curve. However, it’s important to call out that if the system is under maximum load like during a benchmark like FurMark, the fans are going to spin up to maximum RPM. Fan curve updates mostly impact day to day usage in things like browsing or productivity.
am going for this laptop as my side personal working laptop as my current MSI GT75 is very heavy cant carry it with me its more like desktop , so with FW 16 i will be opening a lot of chrome tabs lets hope it will have no problem , in my current msi i have 1000 tabs open lol 870 on chrome and 200 on brave . My MSI fan not make any noise with 1000 tabs its under 50 decidable. finger crossed for FW16
Surely Chrome unloads tabs that haven’t been used recently. I know Firefox does, or else the laptop I’m on at the moment would have melted long ago. It has nowhere near the specs of a FW16. Yet, the fan is normally not even running, despite 3955 tabs across 3 windows. Plus a second Firefox profile running with a few more tabs.
LOL 3955 tabs , i thought its just me who have this bad habit of opening tabs R&D and leave it for later . LOL you break my record . i have 1000 tabs opened in series of 9 chrome windows . 3955 tabs in 3 windows will be quite difficult to switch through lol
On notebookcheck ThinkPad Z16 Gen2 have better (smaller) temperatures where You usually place your hands while typiing (both on Idle and on Max load), than Framework 16.
I really hoped that large laptop as FW16 is, with such an “innovative” cooling system, would do better with the heat dissipation. But after all, I will not cancel my preorder, even though I had to (as a resident of Poland) do some “tricks” to get the order placed.
It depends on the laptop model and how much power they feed the GPU. For the RTX 4060 models coming out ahead of the RX 7700S, they’re simply feeding it more power, up to 140W I believe.
