So, because your processor can only fit 512 instructions, and because it's a RISC processor, it's feasible to either write your own compiler, use colorforth, write it in straight assembly, or some combination thereof. Compilers are pretty bad at generating succinct code, and OK at generating fast code, and a RISC processor might just fit inside your head.

So, I really see two applications for this product: first, to create a supercomputer for fixed-point uses (maybe crypto) with 1 "petaFLOP" (I know, we're not actually talking floating-point multiplications here) for a fraction of the price and much greater energy efficiency. It'd be a world-class supercomputer for $200,000.

The other application would be to embed it in mobile devices to take advantage of its energy efficiency.

What do you imagine the language would be to program that supercomputer?

A Lisp that goes straight to asm. Oh yeah.

What Lisp do you have in mind? Presumably you mean something more specific by "goes straight to asm" than just "is compiled".

If I were going to do this I would build a Lisp bottom-up from the instruction set and my goal would be to find the most interesting language could emerge that way. On the other hand, that's so similar to what Forth already does that it might be better just to treat the hardware as the Forth machine it already is and do what it wants.

You'd take the Richard Gabriel approach, in which you do all kinds of tricks by playing around with macros and types.

Of course you'd probably run it in X86 and let that compile to a binary you propagate through your supercomputer.