G-NMP: Accelerating Graph Neural Networks with DIMM-based Near-Memory Processing.
Teng TianXiaotian WangLetian ZhaoWei WuXuecang ZhangFangmin LuTianqi WangXi JinPublished in: J. Syst. Archit. (2022)
Keyphrases
- neural network
- auto associative
- memory requirements
- artificial neural networks
- pattern recognition
- fuzzy logic
- random access
- real time
- weighted graph
- processing elements
- computational power
- random walk
- connected components
- graph mining
- directed acyclic graph
- fuzzy systems
- data processing
- graph representation
- graph theoretic
- graph model
- memory usage
- social networks
- graph matching
- feed forward
- graph search
- genetic algorithm
- graph based algorithm
- directed graph
- compute intensive
- activation function
- graph partitioning
- training process
- multi layer
- back propagation
- graph theory
- associative memory
- neural network model
- main memory