1: \begin{abstract}

2: Recently Ahmadi et al. (2021) and Tagliaferro (2022) proposed some iterative methods

3: for the numerical solution of linear systems which, under the classical hypothesis of

4: strict diagonal dominance, typically converge faster than the Jacobi method, but

5: slower than the forward/backward Gauss-Seidel one.

6: In this paper we introduce a general class of iterative methods, based on

7: suitable splittings of the matrix that defines the system, which include all of

8: the methods mentioned above and have the same cost per iteration in

9: a sequential computation environment.

10: We also introduce a partial order relation in the set of the splittings and, partly

11: theoretically and partly on the basis

12: of a certain number of examples, we show that such partial order is typically

13: connected to the speed of convergence of the corresponding methods.

14: We pay particular attention to the case of linear systems for which the Jacobi iteration

15: matrix is nonnegative, in which case we give a rigorous

16: proof of the correspondence between the partial order relation and the magnitude of

17: the spectral radius of the iteration matrices.

18: Within the considered general class, some new specific promising methods are proposed

19: as well.

20: \end{abstract}

21: