Historically, geophysical datasets have been becoming larger while state-of-the-art processing techniques have simultaneously become significantly more complex. Large scale computation is now an indispensible component of geophysics and our group seeks to exploit advances in numerical methods and parallel architectures to solve geophysical problems in fundamentally new ways. We are actively adapting our wavefield modeling and tomographic imaging codes to function on PC clusters using domain decomposition techniques. Another area of active research is application of parallel monte-carlo type search techniques for non-linear inversion and error estimation. We currently maintain a 16 node Athlon cluster with 1/2 terrabyte of disk for production scale modeling runs and testing of parallel algorithms.