Revisiting recursive methods for Dyson and Keldysh in NEGF: Part I

TL;DR

This paper reformulates the Recursive Green's Function method using domain decomposition and Schur complement theory to enable parallel quantum transport simulations on modern multi-core clusters.

Contribution

It extends RGF to block n-diagonal systems and introduces a parallel algorithm, DDRGF, for high-performance quantum transport simulations.

Findings

Validated algorithms with Julia implementation showing robustness and scalability.

Extended RGF to handle higher-order stencils and parallel execution.

Provided a clear, reproducible formulation for block-sparse structures in quantum transport.

Abstract

The simulation of quantum transport in nanodevices requires the solution of the Dyson and Keldysh equations, a task dominated by the inversion of massive, block-tridiagonal matrices. While the Recursive Green's Function (RGF) method has long been the standard $O(N)$ solver for quasi-1D systems, its formulation has typically been restricted to sequential execution and nearest-neighbor interactions. In this work, we carefully reformulate RGF through the lens of Domain Decomposition and Schur Complement theory. This allows us to extend the recursive formalism to block $n$-diagonal systems (handling higher-order stencils) and to derive a parallel algorithm, Domain-Decomposition based RGF (DDRGF), which stitches macroscopic domains via reduced interface systems. We explore data dependencies in DDRGF in detail, by means of block-sparse structures and tracing back to the desired output as a block tridiagonal approximation, giving a clear, reproducible and extensible formulation. We validate these algorithms using \texttt{LibNEGF.jl}, a Julia-based implementation, demonstrating that the structural insights of domain decomposition provide a robust pathway for high-performance quantum transport simulations on modern multi-core clusters. The theory presented here lays down the base for tackling the Keldysh problem, to be similarly handled in future stages of our work. Although the target here is the acceleration of kernels in the non-equilibrium Green's function method, the algorithms and the implementations presented can be immediately used in any application involving block $n$-diagonal systems.