Dear all, Some days ago I asked a question [0] here regarding performance of Sage and fast graph algorithms.
For a project I am/will be working on, I am thinking about *adding a interface/library for homomorphism counting into Sage*, which should be helpful for many researchers. This implementation is largely written in C++ (including a little CUDA for parallel computation), with some Java code for tree decomposition (from the PACE 2017 challenge). In the Sage developer documentation, it's only written for using external libraries [1] and packing third-party codes [2]. There are no results for "Java" or "GPU". My questions: *1) What would be a good approach to integrate C++ codes into Sage, via Cython? * *2) In the Sage repo on GitHub, there is no Java files. Is it possible to integrate Java into Sage as well, or it would involve some JVM/GraalVM issues?* *3) Homomorphism counting can be slow on very large graphs, so it would helpful to use GPU for computation. My plan is to use CPU for computing by default, and allow users to enable GPU computing via an optional argument.* *I couldn't find anything on GitHub Sage repo or the developer documentation, so I was wondering are GPU and CUDA currently supported by Sage? (In [0], cugraph [3] was also mentioned) If so, what would be the recommended approaches here?* Many thanks. [0]: https://groups.google.com/g/sage-devel/c/c7wZIrMGqIE [1]: https://doc.sagemath.org/html/en/developer/coding_in_other.html [2]: https://doc.sagemath.org/html/en/developer/packaging.html [3]: https://github.com/rapidsai/cugraph Jing -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/db20b872-68f8-4543-843f-25364f73584bn%40googlegroups.com.
