Enabling energy-efficient DNN training on hybrid GPU-FPGA accelerators.
Xin HeJiawen LiuZhen XieHao ChenGuoyang ChenWeifeng ZhangDong LiPublished in: ICS (2021)
Keyphrases
- energy efficient
- wireless sensor networks
- training process
- field programmable gate array
- energy consumption
- sensor networks
- real time
- data dissemination
- single chip
- routing protocol
- graphics processing units
- parallel computing
- base station
- data gathering
- energy efficiency
- multi core architecture
- high speed
- sensor nodes
- computing systems
- hardware implementation
- hardware architecture
- low cost
- data transmission
- embedded systems
- ad hoc networks
- data acquisition
- cost effective