|
The FEL program has a 64 node 128 processor cluster that is designed to solve FEL based models in 4 dimensions i.e. x, y, z, time. This system utilizes UNIX, OS X, FFTW, LAM MPI, and custom designed C code developed by Professor Colson, and Professor Blau to solve FEL related problems in a fraction of the time that was previously required. The cluster has a theoretical performance of 0.5 Terra Flops. When it is not performing FEL calculations it also is used as an animation render farm.
Picture Of the Cluster under development.
|
