1: \begin{abstract}

2: We consider asynchronous versions of the first and second order Richardson

3: methods for solving linear systems of equations.

4: These methods depend on parameters whose values are chosen {\em a priori}.  

5: We explore the parameter values that can be proven to give convergence

6: of the asynchronous methods.  This is the first such analysis for

7: asynchronous second order methods.  We find that for the first order

8: method, the optimal parameter value for the synchronous case also gives an

9: asynchronously convergent method. For the second order method, the parameter

10: ranges for which we can prove asynchronous convergence do not contain

11: the optimal parameter values for the synchronous iteration.  In practice,

12: however, the asynchronous second order iterations may still converge

13: using the optimal parameter values, 

14: or parameter values close to the optimal ones,

15: despite this result.  We explore this behavior with a multithreaded

16: parallel implementation of the asynchronous methods.

17: \end{abstract}

18: