The calculation of eigenmodes in superconducting cavities treated fully 3-dimensional is problematic on usual high end workstations due to the large amount of memory needed and the large number of floating point operations to be performed. Therefore the present approach uses a parallel SIMD supercomputer (APE-100) to deal with the task of finding the eigenvalues and associated eigenvectors of a large sparse matrix. The matrix is built up by the commercial software tool MAFIA and then sent to the nodes of the supercomputer where the package MAXQ solves the eigenvalue problem. The result of the diagonalization procedure is then read back to the MAFIA host where further data analysis and visualization can be done.
展开▼