Key points are not available for this paper at this time.
The evolution of memristors and their successful applications have positioned them as formidable candidates for the next generation of computer systems. With the rapid advancement of foundational ar- tificial intelligence applications, there is an increasing demand for computational power, energy efficiency, and stability. Memristors and the Neuromorphic Computing (NMC) systems they underpin hold signifi- can’t potential to break through the von Neumann bottleneck. However, technical challenges remain in the application of NMC to computer systems. In this review, we focus on the performance of various structured memristors within Neuromorphic Computing and across different machine learning algorithms. We pro- vide an overview of the current challenges faced by NMC, including the structural limitations due to sneak paths and the inherent power consumption limitations, and offer a perspective on future developments and opportunities in the field.
Jun Zhou (2024) studied this question.