COIN: Communication-Aware In-Memory Acceleration for Graph Convolutional Networks.
Sumit K. MandalGokul KrishnanA. Alper GoksoyGopikrishnan Ravindran NairYu CaoÜmit Y. OgrasPublished in: IEEE J. Emerg. Sel. Topics Circuits Syst. (2022)
Keyphrases
- average degree
- fully connected
- computer networks
- edge weights
- graph structures
- graph representation
- random walk
- social networks
- high bandwidth
- state information
- directed graph
- graph layout
- memory requirements
- dynamic networks
- small world
- communication systems
- bipartite graph
- graph structure
- network structure
- network connectivity
- network analysis
- random graphs
- graph model
- highly connected
- structured data
- community detection
- community discovery
- graph mining
- graph theory
- overlapping communities
- clustering coefficient
- degree distribution
- graph theoretic
- betweenness centrality
- packet switched
- protein interaction networks
- path length
- power law
- credit card
- network architecture
- graph matching
- main memory
- connected components
- complex networks
- shortest path
- semi supervised