parrsb-exascale-spectral-element-mesh

name	parrsb-exascale-spectral-element-mesh
description	parRSB: Exascale Spectral Element Mesh Partitioning (arXiv: 2606.14659v1). We introduce parRSB - a parallel, highly scalable graph partitioner for spectral...

parRSB: Exascale Spectral Element Mesh Partitioning

We introduce parRSB - a parallel, highly scalable graph partitioner for spectral element meshes that produce high quality partitions. parRSB is based on Recursive Spectral Bisection (RSB) algorithm implemented on the dual graph of the input mesh. RSB uses the Fiedler vector, which is the eigenvector

Paper Information

• arXiv ID: 2606.14659v1
• Authors: Summit and Frontier Supercomputer Team
• Published: 2026-06-12
• Category: cs.DC
• Links:
  • Abstract: https://arxiv.org/abs/2606.14659v1
  • PDF: https://arxiv.org/pdf/2606.14659v1

Core Methodology

We introduce parRSB - a parallel, highly scalable graph partitioner for spectral element meshes that produce high quality partitions. parRSB is based on Recursive Spectral Bisection (RSB) algorithm implemented on the dual graph of the input mesh. RSB uses the Fiedler vector, which is the eigenvector associated with the smallest non-zero eigenvalue of the Laplacian matrix of the dual graph for making partitioning decisions and tries to minimize the communication volume between the partitions.

We implemented two numerical methods: Lanczos, and Inverse iteration using Conjugate Gradient method to compute the Fiedler vector. We present partitioning results using parRSB on Summit and Frontier supercomputers at Oak Ridge National Laboratory to illustrate the quality of the partitions produced by parRSB and the scalability of our implementation. We also present results for some of the optimizations we did to speed up the partitioning process.

Key Contributions

1. Novel approach to systems engineering
2. Category: cs.DC
3. Research domain: Distributed Computing

Technical Approach

Problem Domain

We introduce parRSB - a parallel, highly scalable graph partitioner for spectral element meshes that produce high quality partitions. parRSB is based on Recursive Spectral Bisection (RSB) algorithm im

Methodology Highlights

• Systems engineering principles
• Exascale computing
• Verification and optimization patterns

Activation Keywords

• systems engineering
• parrsb-exascale-spectral-element-mesh
• distributed systems
• verification

Related Skills

• systems-engineering-apr2026
• modern-systems-engineering-patterns
• distributed-system-resiliency

Applications

• HPC mesh partitioning
• Systems verification
• Error detection and correction

Limitations & Pitfalls

• Preprint status (not peer-reviewed)
• Specialized domain knowledge may be required
• Check for updated versions on arXiv

Implementation Notes

This skill captures the research methodology from the arXiv paper. Apply the patterns and approaches described in the abstract to relevant systems engineering problems.

Source

arXiv: 2606.14659v1 - parRSB: Exascale Spectral Element Mesh Partitioning