Is there any chance that next RockPi will be released as a SOM right from the production start?
The RockPi format is not really suitable for integration into some other system, but it is possible to make a RockPi SBC with core module and motherboard.
I’m not sure if that will be more expensive to manufacture thought.
I doubt on this as SOM is not sold easily, it is only used for special use cases which means it will sell less.
While an sbc have much larger audience.
Someone from Radxa team can give you the accurate answer to your question.
the thing is that you can use SOM ti actually manufacture an SBC
And the second thing is that SOMs are usually purchased in quantity, and I almost sure that it would outsell SBC easily (if the software support will be OK of course)
While i indeed agree with you that SOM concept is interesting idea to implement it Radxa would need to create SOM and then Carrier Board. And then re-do many thing with SBC. No very optional resource management in theory. Am i wrong?
SOM contains all the high speed components and power supplies, so the SBC board will be rather simple,
And I’m not asking to redo something that already exists - I’m asking about a new development.
It’s still too early to talk about the RK3588. I think for the SoM, it’s not individual user oriented, while SBC is for everyone.
The RockPi format is not really suitable for integration
I think this is the core issue.
rk3588 looks great on paper. I’v been considering the “brain” for my robot project, was thinking about RP4 CM on Gumstix Pixhawk carrier board but not a perfect fit, at that price point I would expect better. If I have to build a custom carrier board, I wouldn’t choose RP4 CM.
I’m not trying to tell that Radxa should sell SOM instead of SBC.
I’m trying to say that with SOM, Radxa can release the very same SBCs (based on these SOMs)
I already posted a photo of Radxa SBC that illustrates this.
here is an example how a good SOM with good support “works”