Ferroelectric FET analog synapse for acceleration of deep neural network training

Physics & Astronomy

• synapses 95%
• education 74%
• analogs 61%
• field effect transistors 59%
• learning 55%
• chips 40%
• harnesses 34%
• electric potential 27%
• polarization 27%
• pulse duration 20%
• histories 20%
• trajectories 19%
• requirements 17%
• energy 14%
• pulses 12%
• performance 11%

Engineering & Materials Science

• Ferroelectric materials 100%
• Field effect transistors 81%
• Deep neural networks 73%
• Data storage equipment 46%
• Polarization 46%
• RRAM 35%
• Dynamic random access storage 29%
• Electric potential 26%
• Macros 22%
• Energy efficiency 19%
• Trajectories 17%
• Networks (circuits) 13%

Chemical Compounds

• Polarization 56%
• Voltage 54%
• Weight 40%
• Conductance 36%
• Energy 35%
• Time 13%