At present the technology is allowing users to take advantage of unused processing power over corporate networks for the execution of large, complex calculations. Pratt & Whitney engineers are running computing tasks for calculations on jet engine designs and materials stress over a grid comprising 8,000 processors on 5,000 workstations in three cities.
Gary Barnett, an analyst with Ovum, said, "Today it belongs to those who want to do large-scale computations but the first steps are being made towards more sophisticated scalable shared computing across the world.
"What has made it possible is the connectivity available with IP. The next evolution will come from developments in clustering, management and monitoring software," he added.
Grid computing software allows resource management and provides a way of viewing machines across the grid to see which are available to do extra processing work and then sharing out work out between them. Developers do not need to know the specifications of each machine, they simply submit the calculation and the software shares the necessary files and data between CPUs.
Supplier roadmaps for grid computing are closely linked with their Web services plans. The draw for users is the potential to be able to sweat their existing assets to carry out tasks normally done by supercomputers.