Hi,

I would like to say something about addressing memory.
2MB large ? Compared to what ? When I started with a microprocessor, I had 256 Bytes of memory that is 0.25 KB. So what is large ?

I saw banking and paging solutions in all those years in many different ways.

Paging was used to get a faster interface to Dynamic Ram. The Row address is the page address and the Column address is used to address within a page. The processor is addressing memory in the usual way. The memory controller hardware addresses a page (Row) and then address within a page (Column) The next read from the same page can be done without the need to address the Page again.
This makes the access to the memory faster within one page. But it will cost some hardware to accomplish this. You need Page ram and (fast) hardware to detect that the current processor address is within (or not) the same page as the previous access.

Banking is used in several ways and I think that is the best way to address large memory. How it is build depends on the application. Two situations:

Just divide the address range in half (or 25/75) Use the lower 32K for program, I/O and stack. Use the upper 32K for addressing within a memory bank.
Use an output register (as many as you need) in the lower address space for the bank addressing.

You can also address the bank with a counter. If you have hardware that can write to memory, you can fill memory banks with data and use counters for addressing. You need extra hardware to access memory and signaling to the processor if hardware is writing.
In both solutions, you can address an ‘infinite’ amount of memory. You only need more hardware to address the banks

There are many more solutions are possible. I think it most depends on the application but the question is pure theoretical as you said.