Jez,

There's a range of ways in which "doesn't work" can be interpreted. The specific case to which I referred in that PIC-related comment I made was that I snagged several version of a 16C5x core and found that not one of my friends, nor I could make 'em execute code that we could get running just fine on a 16C57 (I think that's what it was (it's a 0.6"-wide old original windowed CMOS EPROM PIC) that's seemingly capable of most of what any of the PIC's of that series will do.

No matter what my more experienced-with-VHDL/Verilog-than-I friends tried, they couldn't gat any of these purportedly complete PIC cores to run. Likewise, we've been collaborating on getting some of the published 805x soft-cores to function, and, so far, have not found even one that works as described, though some simply don't have much documentation, so that's not a good measure. The problem is not so much that they don't work as labelled, but, rather, lack of indication on the label.

A couple have worked, ..."sorta..." but not yet enough that we'd try to do useful work with it. Mostly it's problems with the limited documentation or lack of it, or perhaps as you've indicated, the work has been indefinitely suspended for lack of enthusiasm.

As a consequence, I've been fooling with one of my own, a scratch-built version that attempts to accelerate its operation by executing multiple instructions and fetching multiple bytes from the program store at once, thereby potentially enabling it to execute some instructions out of sequence and others concurrently. We'll see where that leads, if anywhere.

RE