pizza / genx

generates functions in native x86 machine code by way of a genetic algorithm

This URL has Read+Write access

genx /
name age message
file LICENSE Fri Feb 27 01:03:45 -0800 2009 eliminated unneeded docs [pizza]
file README Loading commit data...
directory doc/
directory misc/
directory src/
README
# ex: set ff=dos ts=2 et:
# $Id$

git clone git://github.com/pizza/genx.git ~/genx && cd ~/genx/src && make && ./genx

 o=O   
0===0  genx searches for functions via a genetic algorithm, creating them
0===0  randomly from x86 machine language. The "fittest" ones are retained
 O=o   and mutated. Over time, better and better functions evolve.
  O    
 o=O   Why is this interesting?
0===0  
0===0  - Discover Unknown/Unexpected Relationships
 O=o     Given input and output data, genx can search for the simplest
  O      relationship between them.
 o=O   
0===0  - Automated Optimization
0===0    Given input and an existing function, genx can try to find simpler,
 O=o     more efficient equivalences.
  O    
 o=O   - It's Fast
0===0    Because the functions are expressed in native x86 they can be
0===0    executed by the CPU directly, allowing even my modest old laptop to
 O=o     process millions of candidates per minute. 
  O    
 o=O   Requirements
0===0     x86 processor, make, gcc
0===0  
 O=o   Build
  O      make
 o=O   
0===0  Run 
0===0    cd src && ./genx
 O=o   
  O    Enjoy
 o=O