Binary‐Stochasticity‐Enabled Highly Efficient Neuromorphic Deep Learning Achieves Better‐than‐Software Accuracy
Yang Li,
Wei Wang,
Ming Wang
et al.
Abstract:In this work, the requirement of using high‐precision (HP) signals is lifted and the circuits for implementing deep learning algorithms in memristor‐based hardware are simplified. The use of HP signals is required by the backpropagation learning algorithm since the gradient descent learning rule relies on the chain product of partial derivatives. However, it is both challenging and biologically implausible to implement such an HP algorithm in noisy and analog memristor‐based hardware systems. Herein, it is dem… Show more
“…It is further found that the low significant weights are not necessary to participate in the information forwarding and error backpropagation processes of the neural network [51,[62][63][64]. Thus, the gradient of the loss to the weight could be accumulated in a separate array outside of the crossbar of memristive devices, as shown in figure 6(f).…”
Crossbar arrays of memristors are promising to accelerate the deep learning algorithm as a non-von-Neumann architecture, where the computation happens at the location of the memory. The computations are parallelly conducted employing the basic physical laws. However, current research works mainly focus on the offline training of deep neural networks, i.e., only the information forwarding is accelerated by the crossbar arrays. Two other essential operations, i.e., error backpropagation and weight update, are mostly simulated and coordinated by a conventional computer in von Neumann architecture, respectively. Several different in situ learning schemes incorporating error backpropagation and/or weight updates have been proposed and investigated through simulation. Nevertheless, they met the issues of non-ideal synaptic behaviors of the memristors and the complexities of the neural circuits surrounding crossbar arrays. Here we review the difficulties in implementing the error backpropagation and weight update operations for online training or in-memory learning that are adapted to noisy and non-ideal memristors. We hope this work will bridge the gap between the device engineers who are struggling to develop an ideal synaptic device and neural network algorithmists who are assuming that ideal devices are right at hand. The close of this gap could push forward the information processing system paradigm from computing-in-memory to learning-in-memory, aiming at a standalone non-von-Neumann computing system.
“…It is further found that the low significant weights are not necessary to participate in the information forwarding and error backpropagation processes of the neural network [51,[62][63][64]. Thus, the gradient of the loss to the weight could be accumulated in a separate array outside of the crossbar of memristive devices, as shown in figure 6(f).…”
Crossbar arrays of memristors are promising to accelerate the deep learning algorithm as a non-von-Neumann architecture, where the computation happens at the location of the memory. The computations are parallelly conducted employing the basic physical laws. However, current research works mainly focus on the offline training of deep neural networks, i.e., only the information forwarding is accelerated by the crossbar arrays. Two other essential operations, i.e., error backpropagation and weight update, are mostly simulated and coordinated by a conventional computer in von Neumann architecture, respectively. Several different in situ learning schemes incorporating error backpropagation and/or weight updates have been proposed and investigated through simulation. Nevertheless, they met the issues of non-ideal synaptic behaviors of the memristors and the complexities of the neural circuits surrounding crossbar arrays. Here we review the difficulties in implementing the error backpropagation and weight update operations for online training or in-memory learning that are adapted to noisy and non-ideal memristors. We hope this work will bridge the gap between the device engineers who are struggling to develop an ideal synaptic device and neural network algorithmists who are assuming that ideal devices are right at hand. The close of this gap could push forward the information processing system paradigm from computing-in-memory to learning-in-memory, aiming at a standalone non-von-Neumann computing system.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.