Reinforcement learning-based register renaming policy for simultaneous multithreading CPUs.
Huixin ZhanVictor S. ShengWei-Ming LinPublished in: Expert Syst. Appl. (2021)
Keyphrases
- multithreading
- reinforcement learning
- optimal policy
- policy search
- parallel computing
- action selection
- computational power
- markov decision process
- highly efficient
- parallel processing
- policy gradient
- control policy
- state space
- function approximators
- reward function
- temporal difference
- markov decision processes
- distributed memory
- function approximation
- memory efficient
- real time
- coarse grained
- action space
- data partitioning
- multi core processors
- dynamic programming
- reinforcement learning algorithms
- shared memory
- model free
- learning algorithm