Efficient VLSI implementation of iterative solutions to sparse linear systems (Q1801374)

In each step of an iterative method to solve a sparse linear system \(Ax=b\), the major computational step is \(x_ k=Px_{k-1}+q\) with a sparse matrix \(P\) (possibly different from \(A\)). The authors propose a way to perform this step on a regular processor grid with a nearest neighbour communication network.