Releases: youwuyou/Ginkgo.jl
v1.0.0
Ginkgo v1.0.0
1.0.0
We are excited to announce the release of Ginkgo.jl 1.0.0, a milestone in providing Julia users with access to the high-performance numerical linear algebra capabilities of the Ginkgo library.
- Julia Wrapper for Ginkgo. Ginkgo.jl integrates the Julia programming environment with Ginkgo, a high-performance numerical linear algebra library, through a C API.
- Structured and Comprehensive. The package includes a documentation using Documenter.jl, a minimal-working example of a simple solver, and various tests ensuring a complete and user-friendly experience.
- Installation and Use. It can be easily installed using Pkg.jl and used in Julia scripts, making it accessible for users.
- Testing Framework. A testing framework is provided to ensure reliability.
- Feature Implementation. The release includes various implemented, partially implemented, and planned features across different computational paradigms like OpenMP, CUDA, HIP, and DPCPP/SYCL. Key components like Dense and Sparse Matrix CSR, Solvers, and IO are partially implemented, marking progress towards comprehensive functionality.
Getting Started
To learn how to use Ginkgo, and get ideas for your own projects, take a look at the following example:
- simple-solver is an example that reads the matrices from files, computes the final residual, and selects a different executor based on the command-line parameter.
Ginkgo's sources can also serve as a good example, since built-in components are mostly implemented using publicly available utilities.
About
Ginkgo 1.0.0 is a result of a 3-month undergraduate research assistant internship from 09 - 11/2023 conducted by
You Wu, ETH Zürich
under the supervision of
Tobias Ribizel, Karlsruhe Institute of Technology
Valentin Chuvary, Massachusetts Institute of Technology
Hartwig Anzt, Karlsruhe Institute of Technology
at Innovative Computing Laboratory (ICL) in the University of Tennessee
Merged pull requests:
- CompatHelper: add new compat entry for CEnum at version 0.5, (keep existing compat) (#5) (@github-actions[bot])
Closed issues:
- Create a minimal working example using a dummy library to understand how the C API functions (#11)