A 2.1TFLOPS/W Mobile Deep RL Accelerator with Transposable PE Array and Experience Compression.
Changhyeon KimSanghoon KangDongjoo ShinSungpill ChoiYoungwoo KimHoi-Jun YooPublished in: ISSCC (2019)
Keyphrases
- reinforcement learning
- mobile devices
- image compression
- data compression
- mobile phone
- mobile applications
- state space
- user experience
- mobile computing
- mobile environments
- multi agent
- context aware
- markov decision processes
- compression scheme
- mobile networks
- compression ratio
- smart phones
- autonomous learning
- neural network
- embedded systems
- mobile technologies
- computing environments
- wavelet transform
- model free
- learning algorithm
- machine learning