cudaverse: Lightweight 'CUDA' Numerical Computing

Provides a lightweight interface to graphics processing unit (GPU)-accelerated numerical computing using 'CUDA'. Dense tensors, sparse matrices, decompositions, distances, exact nearest neighbours, clustering, graph workflows, and embeddings share one consistent interface. The native backend discovers the 'NVIDIA CUDA Driver API', 'cuBLAS', and 'cuSOLVER' libraries at runtime without bundling 'LibTorch' or the 'CUDA Runtime'. Stage-level provenance records the backend, device, and data transfers used by each result. A portable implementation supports package validation on systems without 'CUDA'. Background for the included Leiden community detection and uniform manifold approximation and projection methods is given by Traag, Waltman and van Eck (2019) <doi:10.1038/s41598-019-41695-z> and McInnes et al. (2018) <doi:10.21105/joss.00861>, respectively.

Version: 0.4.1
Imports: Matrix, methods, stats
Suggests: igraph (≥ 2.0.0), knitr, rmarkdown, RSpectra, Rtsne, S4Vectors, SingleCellExperiment, testthat (≥ 3.0.0), torch, uwot
Published: 2026-09-10
DOI: 10.32614/CRAN.package.cudaverse (may not be active yet)
Author: Yaoxiang Li [aut, cre]
Maintainer: Yaoxiang Li <liyaoxiang at outlook.com>
BugReports: https://github.com/cudaverse/cudaverse/issues
License: MIT + file LICENSE
URL: https://cudaverse.github.io/cudaverse/, https://github.com/cudaverse/cudaverse
NeedsCompilation: yes
SystemRequirements: For GPU execution on Windows or Linux: NVIDIA CUDA-capable GPU, NVIDIA driver with CUDA Driver API, NVIDIA cuBLAS 12, and NVIDIA cuSOLVER 11
Language: en-US
Materials: README, NEWS
CRAN checks: cudaverse results

Documentation:

Reference manual: cudaverse.html , cudaverse.pdf
Vignettes: Keep an R workflow on the GPU (source, R code)
What runs on CUDA? (source, R code)
Your first CUDA workflow with cudaverse (source, R code)
Set up CUDA for cudaverse (source, R code)
Performance and package size (source, R code)

Downloads:

Package source: cudaverse_0.4.1.tar.gz
Windows binaries: r-devel: not available, r-release: not available, r-oldrel: not available
macOS binaries: r-release (arm64): not available, r-oldrel (arm64): not available, r-release (x86_64): not available, r-oldrel (x86_64): not available

Linking:

Please use the canonical form https://CRAN.R-project.org/package=cudaverse to link to this page.