Review of Cyclic Reduction for Parallel Solution of Hermitian Positive Definite Block-Tridiagonal Linear Systems
Martin Neuenhofen

TL;DR
This paper reviews a numerically stable, parallel-friendly cyclic reduction method tailored for Hermitian positive definite banded linear systems, highlighting its advantages and robustness in determining positive definiteness.
Contribution
It provides a comprehensive review of a specialized cyclic reduction method optimized for Hermitian positive definite systems, emphasizing its stability and parallel efficiency.
Findings
Method is numerically stable without pivoting
Suitable for parallel computations
Robust in determining positive definiteness
Abstract
Cyclic reduction is a method for the solution of (block-)tridiagonal linear systems. In this note we review the method tailored to hermitian positive definite banded linear systems. The reviewed method has the following advantages: It is numerically stable without pivoting. It is suitable for parallel computations. In the presented form, it uses fewer computations by exploiting symmetry. Like Cholesky, the reviewed method breaks down when the matrix is not positive definite, offering a robust way for determining positive definiteness.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsMatrix Theory and Algorithms · Advanced Numerical Analysis Techniques · Numerical methods for differential equations
