NetPU-M: a Generic Reconfigurable Neural Network Accelerator Architecture for MLPs.
Yuhao LiuShubham RaiSalim UllahAkash KumarPublished in: IPDPS Workshops (2023)
Keyphrases
- neural network
- multilayer perceptron
- multi layer perceptron
- network architecture
- neural architecture
- hidden neurons
- hardware implementation
- activation function
- low cost
- associative memory
- domain specific
- software architecture
- systolic array
- field programmable gate array
- artificial neural networks
- management system
- back propagation
- self organizing maps
- compute intensive
- heterogeneous computing
- pattern recognition
- radial basis function
- neural network model
- multi layer
- real time
- reconfigurable architecture
- feed forward neural networks
- feed forward
- hardware architecture
- prediction model
- genetic algorithm
- dynamic reconfiguration
- general purpose
- training process
- feedforward neural networks
- fuzzy neural network
- data flow
- fuzzy logic