Applications.

Forced volunteer computing can be used wherever there is a need or opportunity to pool existing (and idle) resources to attain supercomputing power that would otherwise be unaffordable. For example, companies can use computers on their intranet for solving CPU-intensive problems such as market and process simulations, and data-mining [12]. Similarly, research labs and universities can use volunteer computing to turn their existing networks of workstations into virtual supercomputers. In financially-constrained institutions such as universities in developing countries, volunteer networks can also be used to teach supercomputing techniques, without requiring actual (expensive) supercomputers.

By making it very easy for institutions around the world to share their computing resources, volunteer computing opens up exciting new possibilities in world-wide cooperative and collaborative research efforts. For example, research institutions on opposite sides of the globe could barter-trade for each other's computing power (barter trade can also be a form of paid volunteer computing, see Sect.3.3) - a university in the United States can allow a lab in Japan use of its CPUs at night, when it is day in Japan, in exchange for being able to use the Japanese lab's CPUs during the day, when it is night in Japan.

Forced volunteer computing is already being done by many institutions using tools such as PVM or their own ad hoc tools ([13] has some links to PVM-based projects). Java-based forced volunteer computing improves on these by being much easier to use for everyone - users, programmers, and administrators. If a company decides to use its machines to do parallel computation, for example, the administrators would not need to spend time manually installing the computational software to be run on all the company machines - they could simply tell their employees to point their browsers to a certain web page on the company intranet, and leave the browser running while they work, or when they go home. In this way, setting-up a parallel computation, a process that would normally take weeks for software installation and user education, can be done literally overnight.