On Thu, 22 May 2003, Dwight K. Elvey wrote:
> >From: "ben franchuk" <bfranchuk_at_jetnet.ab.ca>
> >
> >David Comley wrote:
> >
> >> Since you mention it, I had been thinking about
> >> designing and building a processor from scratch from
> >> TTL devices. I am slowly accumulating TTL chips as I
> >> come across them at hamfests and things. Perhaps it's
> >> time to put pencil to paper.
> >>
> >> Of course I could take the NASA Apollo Guidance
> >> Computer approach and build everything welded-cordwood
> >> style out of NOR gates.
> >
> >Nope that used lots of REAL ( expensive ) TTL.
> >The neat part of that was the CORE memory used.
> >
> >Any how a real TTL computer is about 4+ large
> >logic cards. Control card, alu card, memory card
> >and serial I/O card. The mother board is bus
> >and front pannel logic.
> >I am doing a 20 bit CPU with about 125 chips total
> >in the computer and front panel. About 50?
> >more chips for memory and serial i/o.
> >
> >http://www.jetnet.ab.ca/users/bfranchuk/ldp/ldp1.html
> >Ben.
> >
>
> Hi
> All make the assumption that you must make 8 bit/16bit
> or, in your case, 20 bit. One can make a 1 bit alu that
> can have data width controlled by instruction. It may
> not be fast at math but much processor time is consumed
> just looking at true/false. If your model doesn't require
> passing data through the alu for mem/mem and mem/io moves,
> a single bitter makes sense.
> ( My Nicolet is a 20 bit machine. )
> Dwight
>
I have considered making a 1 bit stack oriented machine with 4 bit
instructions, long and short jumps and memory reference are done by pushing
the target address a bit at a time onto the pointer stack, then issuing the
jump or memory ref inst (that looks at the 1 bit pointer stack depthwise)...
Peter Wallace
Received on Thu May 22 2003 - 23:03:00 BST