FICHA · AUR

costa

Communication-optimal shuffle and transpose algorithms.

  • Library
  • Dev
  • SCIENCE
  • Dependency only
official+codex · reviewed · May 31, 2026 description in en

Description

Distributed shuffle and transpose operations can be implemented with communication-optimized algorithms for high-performance computing. It is useful for researchers and developers working with large distributed arrays or matrices.

HPC communication libraries depend on MPI, hardware topology, and workload shape. Validate performance and correctness on the target cluster.

Permissions

Permissions not analysed for this source yet.