What It's Used For
Evolutionary Computation

Whether it's used for design, search or optimization, evolutionary computation is one of the "killer grid apps."

For its robust effectiveness and wide applicability, evolutionary computation is a powerful paradigm for solving tough problems - even problems that would normally require a human engineer.

But, computing capacity has been the limiting factor - until now.

The Frontier Grid Platform is perfect for even the largest evolutionary jobs - either on large populations or in domains with long evaluation times. Plus, Parabon's new OriginTM Evolutionary SDK enables you to apply evolutionary algorithms at a scale never before possible, making it possible to employ thousands of caps of capacity to solve problems otherwise unapproachable.

Parabon, its partners and customers launch evolutionary computation jobs across Frontier every day to solve a wide variety of problems including automated antenna design, DNA nanostructure optimization, data mining, decision support, and adaptive network design, just to name a few.

Benefits of Frontier and Origin Combined:

  • Ease of use. Simply add your own domain simulator and let Origin do the rest.
  • Automatic scalability. Execute on thousands of nodes simultaneously without development.
  • Platform independence. Origin is 100% pure Java and runs on practically any platform. Moreover, it supports native binaries, making it easy to incorporate legacy code into your EC application.
  • Extensibility. If Origin doesn't already support your EC algorithm of choice, easily add such with minimal integration.

A wide variety of EC algorithms are supported:

  • Genetic algorithms
  • Evolutionary strategies
  • Genetic programming
  • Opportunistic evolution

Several Paradigms are supported:

  • Master/slave
  • Island model
  • Asynchronous steady state

What is Evolutionary Computation?

Learn what wikipedia has to say.

Case Studies