Accelerating Graph Convolutional Networks Using Crossbar-based Processing-In-Memory Architectures.
Yu HuangLong ZhengPengcheng YaoQinggang WangXiaofei LiaoHai JinJingling XuePublished in: HPCA (2022)
Keyphrases
- memory management
- graph theory
- data processing
- social graphs
- fully connected
- directed graph
- real time
- graph representation
- small world
- random access
- complex networks
- community discovery
- average degree
- computational power
- graph structures
- deep learning
- dynamic networks
- edge weights
- digital signal processors
- random walk
- memory requirements
- graph model
- network size
- response time
- directed edges
- operating system
- citation networks
- information processing
- social network analysis
- connected components
- functional modules
- scheduling algorithm
- bipartite graph
- graph structure
- network analysis
- community detection
- biological networks