I’m here to also confirm that the Crucial 5600MT/s RAM kit CT2K16G56C46S5 runs flawlessly in my Framework 13 7840U with the most recent 3.03 BIOS.
New Batch 5 Framework 13 AMD with Kingston Fury KF556S40IBK2-64 DDR5 5600MT/s CL40-40-40 on Linux 22.04.3 OEM working fine!
Thanks to all the contributors in this thread, it was a great resource.
I can confirm the kingston Fury impact 5600 cl40 (KF556S40IBK2, double sided, dual rank) work just fine, I can also confirm that the crucial 4800 CL40 (CT16G48C40S5 single sided single rank) works fine too (I got that stick really cheap and was curious if frequency had as big an impact as on the ddr4 intel boards) as do combinations of the two.
I did some idle power tests using combinations of the 2, the tests were done with display at min brightness, keyboard backlight, camera, microphone, wifi and bluetooth off and as little as possible running in the background (basically just 2 terminal windows) taking the powerstat readings after it had settled. I have not done much other power tuning, I just got the thing so those values are mostly interesting relative to each other and not in an absolute sense.
- Single 4800 (Crucial 4800 16GB) → 2.79W
- Dual 4800 (Crucial 4800 16GB + Kingston 5600 32GB) → 3.07
- Single 5600 (Kingston 5600 32GB) → 2.83
- Dual 5600 (Kingston 5600 32GB x2) → 3.05W
Interestingly kind of the opposite behavior of the intel, frequency seems to matter quite little (like margin of error little) while channels do a bit. Halving your memory bandwidth to save an extra 0.2W is probably only for the most extreme hyper-milers among us XD.