AoCStream: All-on-Chip CNN Accelerator with Stream-Based Line-Buffer Architecture and Accelerator-Aware Pruning.
Hyeong-Ju KangByung-Do YangPublished in: Sensors (2023)
Keyphrases
- parallel implementation
- real time
- data streams
- high speed
- host computer
- compute intensive
- search space
- low cost
- analog vlsi
- management system
- convolutional neural network
- vlsi implementation
- pruning method
- software architecture
- neural network
- memory access
- pruning algorithm
- cellular neural networks
- associative memory