Storing Cookies (See : http://ec.europa.eu/ipg/basics/legal/cookies/index_en.htm ) help us to bring you our services at overunity.com . If you use this website and our services you declare yourself okay with using cookies .More Infos here:
https://overunity.com/5553/privacy-policy/
If you do not agree with storing cookies, please LEAVE this website now. From the 25th of May 2018, every existing user has to accept the GDPR agreement at first login. If a user is unwilling to accept the GDPR, he should email us and request to erase his account. Many thanks for your understanding

User Menu

Custom Search

Author Topic: using evolution to 'design' circuits  (Read 4226 times)

WilbyInebriated

  • Hero Member
  • *****
  • Posts: 3141
using evolution to 'design' circuits
« on: May 05, 2011, 06:07:47 PM »
the gist of 'evolvable' hardware is simple enough. the configuration bits that program the wiring and circuitry of the fpga (field programmable gate array) become the 'chromosomes' of the individuals to undergo the trial of survival of the fittest. the chip is configured, then set to do some task (like applying various settings at the gate, or something else), and its performance is measured. then the bit streams representing the best-performing configurations are mated together, mutations are added, and new individuals are tested, measured, and either discarded or mated. eventually, a configuration should emerge that’s very good at accomplishing a specific task, albeit might take thousands of generations to appear.

the beauty of it is, that you can use the fpga's as blank evolutionary 'slates' and let evolution fiddle around with the fine details. ie: you can create without any preconceptions built in. so, it’s not told anything about what is 'good' and what is 'bad' or how it achieves the behavior. evolution just plays around making changes, and if the changes produce an improvement, then fine. it doesn’t matter whether it’s changing the circuit design or using just about any weird, subtle bit of physics (note: this doesn't work on the simulation fpgas, only hardware fpgas) that might be going on. the only thing that matters to evolution is the overall behavior. this means you can explore all kinds of ways of building things that are completely beyond the scope of conventional methods. allow evolution to write all the design rules.

FittySense

  • Newbie
  • *
  • Posts: 23
Re: using evolution to 'design' circuits
« Reply #1 on: May 05, 2011, 06:47:34 PM »
Per the norm, you are too late. It was invented by two Canadian researchers working for IBM in the 70's named Mom and Dad. They named their creation IST - (Innovation Station Team)

Unfortunately, it turned passive aggressive and developed a predilection for smiley faces, magnets and do nothing circuits. It was later discovered that the entropy of "doing nothing" is the natural evolutionary path and explains the behavior of old people such as yourself.

Hope that helps!

WilbyInebriated

  • Hero Member
  • *****
  • Posts: 3141
Re: using evolution to 'design' circuits
« Reply #2 on: May 05, 2011, 11:39:51 PM »
Per the norm, you are too late. It was invented by two Canadian researchers working for IBM in the 70's named Mom and Dad. They named their creation IST - (Innovation Station Team)

Unfortunately, it turned passive aggressive and developed a predilection for smiley faces, magnets and do nothing circuits. It was later discovered that the entropy of "doing nothing" is the natural evolutionary path and explains the behavior of old people such as yourself.

Hope that helps!
yet another duarktoo alias to add to the ignore list... thanks for outing yourself troll. ::)