The GPU as Numerical Simulation Engine

Many computer graphics applications require high-intensity numer- ical simulation. The question arises whether such computations can be performed efficiently on the GPU, which has emerged as a full function streaming processor with high floating point perfor- mance. We show in this paper that this is indeed the case using two basic, broadly useful, computational kernels as examples. The first is a sparse matrix conjugate gradient solverand the second a regular-grid multigrid solver. Many realtime applications rang- ing from mesh smoothing and parameterization to fluid solvers and solid mechanics can greatly benefit from these as we demonstrate with a prototype implementation on NVIDIA's GeForce FX, using geometric flow and fluid simulation as application examples.