Optimizations for the Computation of Radiative Corrections

Abstract

Two methods are presented with which the CPU time spent on the calculation of radiative corrections can be significantly reduced. The first is the parallelization of the program, which can be surprisingly simple to implement under certain circumstances often met in the calculation of radiative corrections. The second is the efficient direct calculation of fermion chains. The latter not only improves the overall performance of the program, but introduces better conceptual clarity as well, as it allows for a homogeneous treatment of bosonic and fermionic amplitudes.