Emerging Trends in Memristor-based Computing and Optimization

The field of memristor-based computing and optimization is experiencing a significant shift towards innovative and advanced techniques. Researchers are exploring new ways to characterize and utilize memristors, leading to breakthroughs in areas such as neuromorphic computing and optimization algorithms. The development of novel modeling approaches, such as Gaussian Process Regression and active learning, is enabling more efficient and accurate characterization of memristor-based devices. Furthermore, the application of neuromorphic computing to optimization problems is yielding promising results, with potential benefits including low power, low latency, and small footprint. Noteworthy papers in this area include:

  • A paper introducing a novel Gaussian Process Regression model with active learning for composite current source characterization, achieving significant improvements in accuracy and efficiency.
  • A study on neuromorphic-based metaheuristics, which proposes a new generation of low power, low latency, and small footprint optimization algorithms.
  • A paper presenting a bi-level metaheuristic charging scheme for minimizing energy depletion in wireless rechargeable sensor networks, demonstrating improved performance through a partial charging approach.

Sources

Cell Library Characterization for Composite Current Source Models Based on Gaussian Process Regression and Active Learning

Concept of a System-on-Chip Research Platform Benchmarking Interaction of Memristor-based Bio-inspired Computing Paradigms

State Characterisation of Self-Directed Channel Memristive Devices

Neuromorphic-based metaheuristics: A new generation of low power, low latency and small footprint optimization algorithms

Minimizing the energy depletion in wireless rechargeable sensor networks using bi-level metaheuristic charging schemes

Built with on top of