Good morning All, Hope you are doing well.
i have an idea and i want to know if it is applicable or if i will face problems, And i need your recommendations.
currently i am buying NXP I.MX8QM MEK board, and my project idea is to use local LLM on the board, For that i will need AI accelerator.
my current setup is using RPI5 and AI Hat +2 and i want to level up the project so i will use NXP i.mx8QM board, And i want to use the AI Hat +2 with the NXP board
the AI Hat +2 is Hailo-10 H AI acceleration module and it is interfaced with pcie fpc 16 way connector as follows its signal is standard PCIe signals
The following Image shows it.
So i want to use it with the NXP board and i thought to create a custom M.2 Card that just routes the PCIe signals from the Hat to the M.2 like the following image.
this is just a simple example of the board as a explanation.
so my question, Is what i am thinking of is applicable or there is something i didn't take care of it?
also if this is applicable do i need a special driver or the linux driver of the hat is enough?
your answer will help me a lot and thank you in advance.
