GMI-DRL: Empowering Multi-GPU Deep Reinforcement Learning with GPU Spatial Multiplexing.
Yuke WangBoyuan FengZheng WangTong GengAng LiYufei DingPublished in: CoRR (2022)
Keyphrases
- reinforcement learning
- real time
- graphics hardware
- parallel implementation
- gpu accelerated
- graphics processors
- parallel computation
- graphics processing units
- spatial information
- gpu implementation
- spatio temporal
- parallel processing
- spatial and temporal
- spatial databases
- function approximation
- spatial data
- general purpose
- spatial structure
- cluster of workstations
- image based rendering
- spatial reasoning
- geographic information systems
- markov decision processes
- transfer learning
- optimal policy
- learning algorithm