Reproduce “Accelerating Direct Sparse Solvers with Graph Computing Accelerators” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Jiya Su, François-Henry Rouet, Jeff Beisheim, Robert Lucas, Yuqing Wang, Alexander Fell, Chris Mccarthy, David F. Gleich and 1 more
UpDownMETIS accelerates sparse matrix reordering by implementing the established METIS nested-dissection algorithm on hardware designed for irregular graph work. The system exploits parallelism across graph vertices and recursively created subgraphs. It reports 4–7× shorter reordering runtimes than ParMETIS for medium and large finite-element problems. These results are modeled from simulation and emulation, not measurements on completed UpDown silicon.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.