FICHA · AUR

cusparselt

A high-performance CUDA library dedicated to general matrix-matrix operations in which at least one operand is a sparse matrix

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

Description

Sparse matrix-matrix operations can be accelerated on NVIDIA GPUs with a specialized CUDA library. This is useful for machine learning, scientific computing, and high-performance workloads that use structured sparsity.

It depends on compatible NVIDIA hardware, drivers, CUDA versions, and licensing. Version mismatches can break builds or runtime behavior, so match it to the application stack.

Permissions

Permissions not analysed for this source yet.