Search Results for "numerics"

Fluid2D allows to study a wide variety of 2D flows. It is written entirely in Python. It is both a teaching code and a research code, capable of running from one core to thousands. Its numerics has been chosen to yield to as small dissipation as possible allowing to simulate easily high Reynolds flows, much higher than any of its concurrents. High performances are achieved by writting most of the operations as matrix-vector multiplications, handled by numpy that itself relies on the highly optimized BLAS.