NeuraChip: Accelerating GNN Computations with a Hash-based Decoupled Spatial Accelerator

arXiv:2404.15510v1 Announce Type: cross
Abstract: Graph Neural Networks (GNNs) are emerging as a formidable tool for processing non-euclidean data across various domains, ranging from social network analysis to bioinformatics. Despite their effectiveness, their adoption has not been pervasive because of scalability challenges associated with large-scale graph datasets, particularly when leveraging message passing.
To tackle these challenges, we introduce NeuraChip, a novel GNN spatial accelerator based on Gustavson's algorithm. NeuraChip decouples the multiplication and addition computations in sparse matrix multiplication. This …

abstract accelerator adoption analysis arxiv bioinformatics challenges cs.ar cs.dc cs.lg cs.ne data datasets domains gnn gnns graph graph neural networks hash network networks neural networks non-euclidean processing scalability scale social spatial tool type