Reconfigurable AI chip dynamically adjusts operational speed over 100-fold range and its energy consumption based on workload demands
https://www.eurekalert.org/news-releases/1139132 "memtransistor of self-assembled block polymeric thin films mimics biological neural plasticity... optimizes hardware performance diverse time-series data without requiring separate specialized processors... handles fast signal processing/ slow memory consolidation on same chip dramatically reducing latency/ power consumption... energy-efficient, brain-inspired neuromorphic hardware capable of real-time environmental adaptation... edge AI"