MFxVF computes a float matrix, specified by a parameter, times a float vector, specified by an input.