FPGA acceleration of deep reinforcement learning using on-chip replay management.
Yuan MengChi ZhangViktor K. PrasannaPublished in: CF (2022)
Keyphrases
- reinforcement learning
- high speed
- low cost
- single chip
- low power consumption
- programmable logic
- ibm zenterprise
- field programmable gate array
- real time
- function approximation
- management system
- knowledge management
- information management
- data processing
- data management
- real time image processing
- machine learning
- model free
- high density
- systolic array
- decision making
- information systems
- network on chip
- vlsi implementation
- software implementation
- physical design
- hardware implementation
- low power
- decision support
- state space
- learning process