A generic systolic array building block for neural networks with on-chip learning

Training Deep Convolutional Neural Networks with Resistive Cross-Point Devices

In a previous work we have detailed the requirements for obtaining maximal deep learning performance benefit by implementing fully connected deep neural networks (DNN) in the form of arrays of resistive devices. Here we extend the concept of Resistive Processing Unit (RPU) devices to convolutional neural networks (CNNs). We show how to map the convolutional layers to fully connected RPU arrays such that the parallelism of the hardware can be fully utilized in all three cycles of the backpropagation algorithm. We find that the noise and bound limitations imposed by the analog nature of the computations performed on the arrays significantly affect the training accuracy of the CNNs. Noise and bound management techniques are presented that mitigate these problems without introducing any additional complexity in the analog circuits and that can be addressed by the digital circuits. In addition, we discuss digitally programmable update management and device variability reduction techniques that can be used selectively for some of the layers in a CNN. We show that a combination of all those techniques enables a successful application of the RPU concept for training CNNs. The techniques discussed here are more general and can be applied beyond CNN architectures and therefore enables applicability of the RPU approach to a large class of neural network architectures.

Acceleration of Deep Neural Network Training with Resistive Cross-Point Devices: Design Considerations

In recent years, deep neural networks (DNN) have demonstrated significant business impact in large scale analysis and classification tasks such as speech recognition, visual object detection, pattern extraction, etc. Training of large DNNs, however, is universally considered as time consuming and computationally intensive task that demands datacenter-scale computational resources recruited for many days. Here we propose a concept of resistive processing unit (RPU) devices that can potentially accelerate DNN training by orders of magnitude while using much less power. The proposed RPU device can store and update the weight values locally thus minimizing data movement during training and allowing to fully exploit the locality and the parallelism of the training algorithm. We evaluate the effect of various RPU device features/non-idealities and system parameters on performance in order to derive the device and system level specifications for implementation of an accelerator chip for DNN training in a realistic CMOS-compatible technology. For large DNNs with about 1 billion weights this massively parallel RPU architecture can achieve acceleration factors of 30, 000 × compared to state-of-the-art microprocessors while providing power efficiency of 84, 000 GigaOps∕s∕W. Problems that currently require days of training on a datacenter-size cluster with thousands of machines can be addressed within hours on a single RPU accelerator. A system consisting of a cluster of RPU accelerators will be able to tackle Big Data problems with trillions of parameters that is impossible to address today like, for example, natural speech recognition and translation between all world languages, real-time analytics on large streams of business and scientific data, integration, and analysis of multimodal sensory data flows from a massive number of IoT (Internet of Things) sensors.

Simultaneous Perturbation Learning Rule for Recurrent Neural Networks and Its FPGA Implementation

Recurrent neural networks have interesting properties and can handle dynamic information processing unlike ordinary feedforward neural networks. However, they are generally difficult to use because there is no convenient learning scheme. In this paper, a recursive learning scheme for recurrent neural networks using the simultaneous perturbation method is described. The detailed procedure of the scheme for recurrent neural networks is explained. Unlike ordinary correlation learning, this method is applicable to analog learning and the learning of oscillatory solutions of recurrent neural networks. Moreover, as a typical example of recurrent neural networks, we consider the hardware implementation of Hopfield neural networks using a field-programmable gate array (FPGA). The details of the implementation are described. Two examples of a Hopfield neural network system for analog and oscillatory targets are shown. These results show that the learning scheme proposed here is feasible.