2
$\begingroup$

We are so used to von neumann architecture and say a register machine like the x86. (Also with programming languages built for those machines x86 assembly, C, etc) Is that approach to computing completely separate from a language and system like haskell?

What would a machine look like that was based functional programming paradigms look like? and one not running on a register or stack machine.

This article is relevant to my question:

http://www.stanford.edu/class/cs242/readings/backus.pdf

$\endgroup$
  • 4
    $\begingroup$ This seems too glib to count as an actual answer, but: why wouldn't they look like an x86 running the Haskell environment? Functional programming is a high-level language construct, and the point of high level programming is to insulate you from the messy details of the machine architecture, so why do you think it would be a good idea to remove this insulation? Alternatively, if you really do want both low-level and functional, have you tried looking at combinators? $\endgroup$ – David Eppstein Jan 31 '12 at 0:00
  • 2
    $\begingroup$ Isn't the correct answer to this question simply "No"? $\endgroup$ – Jeffε Jan 31 '12 at 2:31
  • 1
    $\begingroup$ Yeah, not an actual question. $\endgroup$ – beroal Jan 31 '12 at 9:16
  • 1
    $\begingroup$ If I understand your question, the title has nothing to do with the question. Please edit it. $\endgroup$ – Tsuyoshi Ito Jan 31 '12 at 14:28
9
$\begingroup$

Many hardware designs for executing functional programs have been proposed:

and more (lots are cited in the Reduceron papers).

$\endgroup$
2
$\begingroup$

The abstract machine would look like i.e. the untyped lambda calculus or the SKI combinator or just some kind of a LISP dialect.

In hardware it would probably look like a biologic cell, calculating via pattern matching on the genes in the core – directly running that LISP dialect. (You might like to read this blog entry: "New computer language based on Lisp enables biological modeling")

$\endgroup$
  • $\begingroup$ Hey, biology has an efficient garbage collector and arbitrary parallelism… you cannot have that on any imperative machine nor any FPGA. And enzymes are nano-bots, so… it might even be able to compile itself on the hardware-level. But bio-computers are still as experimental as quantum-computers. $\endgroup$ – comonad Jan 31 '12 at 10:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.