Let A be an upper triangular n n nonsingular matrix, and b an n 1 vector. On this example, consider the number of floating point operations needed to solve Ax = b both directly and by using Gauss-Seidel. Is there a significant advantage when using Gauss-Seidel? Your answer should be in terms of float-

ing point operations and you don't need to use explicit constants, just orders of magnitude O() and powers of n.