Showing cs.DCShow all
2 papers · 1 filter
cs.DC2024
SpComm3D: A Framework for Enabling Sparse Communication in 3D Sparse Kernels
Nabil Abubaker, Torsten Hoefler
Existing 3D algorithms for distributed-memory sparse kernels suffer from limited scalability due to reliance on bulk sparsity-agnostic communication. While easier to use, sparsity-…
cs.DC2024
Arrow Matrix Decomposition: A Novel Approach for Communication-Efficient Sparse Matrix Multiplication
Lukas Gianinazzi, Alexandros Nikolaos Ziogas, Langwen Huang +9
We propose a novel approach to iterated sparse matrix dense matrix multiplication, a fundamental computational kernel in scientific computing and graph neural network training. In…