FICHA · MANJARO

rocblas

Next generation BLAS implementation for ROCm platform

  • Library
  • LIBRARY
  • GPU
  • Dependency only
official+codex · reviewed · May 29, 2026 description in en

Description

Provides BLAS routines accelerated for AMD ROCm GPUs, supporting dense linear algebra used by scientific computing, machine learning, and simulation software. It is useful when applications need high-performance matrix and vector operations on AMD hardware.

Performance and correctness depend on compatible ROCm versions, GPU support, and numerical precision. Test workloads before relying on results in production or research.

Permissions

Permissions not analysed for this source yet.