PARALLEL COMPUTATIONAL MICROHYDRODYNAMICS - SCALABLE LOAD-BALANCING STRATEGIES

The hydrodynamic interactions in a suspension of small particles in a viscous fluid is computed by a boundary integral velocity representati on featuring a completed double-layer potential (completed double-laye r boundary integral equation method or CDL = BIEM). A multipole expans ion is used to represent interactions between distance particles, lead ing to a considerable improvement in computational speed. The resultin g large linear system of equations provides an ideal setting for async hronous iterative solvers (block Gauss-Seidel) on a message-passing MI MD parallel computer (Intel iPSC/860 Hypercube) using a supervisor wor kers load-balancing strategy. Our benchmark results as a function of p roblem size and number of processors suggest that our algorithm will s cale successfully to the massively parallel computers of the future.