Abstract:　Contour-integral based methods for interior eigenvalue problems have proven very effective. The most expensive part of these algorithms is the solution of linear systems with large numbers of right hand sides. So far, our BEAST framework has considered one such contour-integral based scheme, FEAST, but we have now extended it to include moment based methods; namely Sakurai-Sugiura methods.  We consider how performance may be improved through the use of moments with a constrained subspace size to reduce the number of right hand sides in these systems, which is a critical parameter for the cost of such integral based schemes. The iterative nature of the resulting solver is well suited to the BEAST framework, allowing us to adjust parameters during computation for further cost reduction. Numerical results show that these algorithmic considerations lead to a noticeable reduction in time for the solver.