[Coco] What would a CoCo successor have to have as a minimum?
jekent at optusnet.com.au
Tue Nov 23 08:52:13 EST 2010
On 23/11/2010 11:31 AM, Theodore (Alex) Evans wrote:
> If one wants to get pedantic, 8ns giving you 125MHz is only 4x faster
> than 25MHz, not 5x faster. Of course it is still more than enough
> headroom for multiplexing the bus between video (accelerated or not)
> and the CPU especially since there is nothing stopping you from giving
> the video side 32 bit access even with no changes to the CPU. The
> memory connection in a CoCo 3 to the GIME is 16-bit enough to get
> nearly the full bandwidth improvement, pure 8-bit access would have
> been barely fast enough for the video side of things alone leaving
> nothing for the CPU.
> If we are modifying the 6x09 CPU, I would like to see support for 24
> or 32 bit addresses in the CPU and a true 16-bit ALU, and yes a wider
> data bus would be nice too. To support all this there would have to
> be another CPU mode because at the very least it would mess up the stack.
Sorry ... I missed your comment.
Yes .. I put generics in System1609 that define the data width and
125MHz = 4 x CPU cycles + 1 x Video access cycle ?
More information about the Coco