52 minutes ago, Geehaf said:
I wonder if it's worth considering the x16 with a full FPGA implementation like the X8? Not sure what it would mean for the YM chip but perhaps there is already FPGA for that too. It would allow for complete "firmware" upgrades in the future and could simplify the overall design. Not sure if it would help reduce manufacturing costs. Yes, my idea is a little radical I know, especially with all the work and time spent so far on the hybrid FPGA / non FPGA chips approach. ?
It's not all that radical, since it's part the original plan. That is the "phase 3", also known as "X16e".
The market for the X16p/X16c exists because of people who WANT the "authentic" ASIC parts. So the X16e is not "instead of" the X16p or X16c, but a lower cost alternative to those boards.
Putting all of it except for a RAM chip inside an FPGA was the plan for the lowest cost member of the family. To do that, it has to be a larger FPGA than the one that is used for Vera in the X16p. AFAIU from the Foenix256 project, an FPGA core for the YM2151 exists, so it seems like it wouldn't be a serious issue including it.
The point of the X8 is that it's not perfectly compatible with the CX16, but it fits in the same FPGA that Vera uses and doesn't need any external RAM, so it would be about half the price of the X16e. So if the X16e costs $90 for the board, the X8 would cost $45 for the board.