Term: Folding@Home

Folding@Home was created by Stanford University to solve immense mathematical and statistical problems without the need of giant supercomputers. Instead Stanford released software where users can voluntary donate their free CPU cycles to compute these problems. Folding@home absorbed the Genome@home in 2004 which expanded the program to protein folding work units. Folding@Home allows the program to crunch millions upon millions bits of data without relying on powerful supercomputers. The primary contributors are hundreds of thousands of personal computer owners who installed the Folding@home client. Many of these contributors join groups in order to compete against other teams for who can compute the most workunits.


Online Computing Community Folding@Home Teams:
[H]ardFolding - http://www.hardfolding.com/
Overclock.net - http://www.overclock.net/f/55/overclock-net-folding-home-team
Folding@EVGA - http://www.evga.com/folding/
OCC Folding@Home - http://www.overclockersclub.com/pages/folding/

These teams will often hold internal competitions (as well as competitions between various teams). This helps promote activity and gets many people involved with donating their extra computing cycles.

Computations involving:
  • energy minimization over Cartesian coordinates, torsional angles or rigid bodies via conjugate gradient, variable metric or a truncated Newton method;
  • molecular, stochastic, and rigid body dynamics with periodic boundaries and control of temperature and pressure;
  • normal mode vibrational analysis;
  • distance geometry including an efficient random pairwise metrization;
  • building protein and nucleic acid structures from sequence;
  • simulated annealing with various cooling protocols;
  • analysis and breakdown of single point potential energies;
  • verification of analytical derivatives of standard and user defined potentials;
  • location of a transition state between two minima;
  • full energy surface search via a "Conformation Scanning" method;
  • free energy calculations via free energy perturbation or weighted histogram analysis;
  • fitting of intermolecular potential parameters to structural and thermodynamic data; and
  • global optimization via energy surface smoothing, including a "Potential Smoothing and Search" (PSS) method.

Future Outlook
On November 18, 2011, Stanford University announced that their popular Folding@home software is now also heading to supercomputers. This will allow for unprecedented access to an enormous amount of processing power. The future of Folding@home is ever changing.


Web Resources:

Related Terminology: Blue Gene, Grid computing, Rosetta@home, Molecular modeling, Genome@home