When Massive GPU Parallelism Ain't Enough: A Novel Hardware Architecture of 2D-LSTM Neural Network.
Vladimir RybalkinNorbert WehnPublished in: FPGA (2020)
Keyphrases
- hardware architecture
- neural network
- processing elements
- associative memory
- recurrent neural networks
- parallel computing
- parallel computation
- field programmable gate array
- parallel processing
- hardware implementation
- hardware architectures
- back propagation
- parallel architectures
- real time
- neural network model
- pattern recognition
- single instruction multiple data
- massively parallel
- data analysis
- artificial neural networks
- shared memory
- parallel implementation
- feed forward
- signal processing
- artificial intelligence
- block matching motion estimation
- information systems
- message passing
- computer vision
- markov random field
- low cost
- genetic algorithm