Can We Trust Undervolting in FPGA-Based Deep Learning Designs at Harsh Conditions?
IEEE Micro（2022）
关键词
power efficiency,application accuracy,FPGA supply voltage,cancause accuracy problems,design-time considerations,correct configuration,target accuracy,autonomous systems,edge computing,harsh environmental conditions,convolutional neural network,comprehensive testing,calibrated infrastructure atcontrolled temperatures,off-the-shelf FPGA,field-programmable gate arrays,awider context,power-efficiency Giga-OP,FPGA-based deep learning designs,harsh conditions,voltage guard-band shift,reliable undervolting designs,temperature -40.0 degC to 50 degC
AI 理解论文
溯源树
样例
生成溯源树，研究论文发展脉络
Chat Paper
正在生成论文摘要