[Coco] FPGA VS Software Emulators

Mark McDougall msmcdoug at iinet.net.au
Wed Jul 26 02:04:25 EDT 2017


On 26/07/2017 3:34 PM, Walter Zambotti wrote:

> How much parallelism can occur in these FPGA devices?

Every logic element in the device effectively runs "in parallel" on each 
clock. It's the transition of your design's "states" and propagation of 
your data through the logic that makes it a sequential logic design, not 
the device itself. So basically, it's between you and the silicon.

> I suppose the down side is the maximum speed of the memory accesses
> will only be half as fast of what they could be!

Again, this comes down more to replicating the original (Coco) design 
and producing an 'emulation' that operates faithfully to the original. 
In theory you could implement instruction/data caches and burst to/from 
(faster) memory for both CPU and video (given appropriate bandwidth 
management) but then you're moving further away from a Coco and start 
having issues with eg. video interrupt timings and you start to break 
software like Sockmaster's demos etc.

Regards,

-- 
Mark McDougall
<http://retroports.blogspot.com.au>


More information about the Coco mailing list