Search for: All records

In-memory computing represents an effective method for modeling complex physical systems that are typically challenging for conventional computing architectures but has been hindered by issues such as reading noise and writing variability that restrict scalability, accuracy, and precision in high-performance computations. We propose and demonstrate a circuit architecture and programming protocol that converts the analog computing result to digital at the last step and enables low-precision analog devices to perform high-precision computing. We use a weighted sum of multiple devices to represent one number, in which subsequently programmed devices are used to compensate for preceding programming errors. With a memristor system-on-chip, we experimentally demonstrate high-precision solutions for multiple scientific computing tasks while maintaining a substantial power efficiency advantage over conventional digital approaches. 

In this work, the effect of rapid thermal annealing (RTA) temperature on the ferroelectric polarization in zirconium-doped hafnium oxide (HZO) was studied. To maximize remnant polarization (2P r ), in-plane tensile stress was induced by tungsten electrodes under optimal RTA temperatures. We observed an increase in 2P r with RTA temperature, likely due to an increased proportion of the polar ferroelectric phase in HZO. The HZO capacitors annealed at 400°C did not exhibit any ferroelectric behavior, whereas the HZO capacitors annealed at 800°C became highly leaky and shorted for voltages above 1 V. On the other hand, annealing at 700 °C produced HZO capacitors with a record-high 2P r of ∼ 64 μ C cm −2  at a relatively high frequency of 111 kHz. These ferroelectric capacitors have also demonstrated impressive endurance and retention characteristics, which will greatly benefit neuromorphic computing applications. 

Asynchronous event-driven computation and communication using spikes facilitate the realization of spiking neural networks (SNN) to be massively parallel, extremely energy efficient and highly robust on specialized neuromorphic hardware. However, the lack of a unified robust learning algorithm limits the SNN to shallow networks with low accuracies. Artificial neural networks (ANN), however, have the backpropagation algorithm which can utilize gradient descent to train networks which are locally robust universal function approximators. But backpropagation algorithm is neither biologically plausible nor neuromorphic implementation friendly because it requires: 1) separate backward and forward passes, 2) differentiable neurons, 3) high-precision propagated errors, 4) coherent copy of weight matrices at feedforward weights and the backward pass, and 5) non-local weight update. Thus, we propose an approximation of the backpropagation algorithm completely with spiking neurons and extend it to a local weight update rule which resembles a biologically plausible learning rule spike-timing-dependent plasticity (STDP). This will enable error propagation through spiking neurons for a more biologically plausible and neuromorphic implementation friendly backpropagation algorithm for SNNs. We test the proposed algorithm on various traditional and non-traditional benchmarks with competitive results. 

In recent years, neuromorphic computing systems (NCS) have gained popularity in accelerating neural network computation because of their high energy efficiency. The known vulnerability of neural networks to adversarial attack, however, raises a severe security concern of NCS. In addition, there are certain application scenarios in which users have limited access to the NCS. In such scenarios, defense technologies that require changing the training methods of the NCS, e.g., adversarial training become impracticable. In this work, we propose AdverQuil – an efficient adversarial detection and alleviation technique for black-box NCS. AdverQuil can identify the adversarial strength of input examples and select the best strategy for NCS to respond to the attack, without changing structure/parameter of the original neural network or its training method. Experimental results show that on MNIST and CIFAR-10 datasets, AdverQuil achieves a high efficiency of 79.5 - 167K image/sec/watt. AdverQuil introduces less than 25% of hardware overhead, and can be combined with various adversarial alleviation techniques to provide a flexible trade-off between hardware cost, energy efficiency and classification accuracy. 

Abstract The increasing interests in analog computing nowadays call for multipurpose analog computing platforms with reconfigurability. The advancement of analog computing, enabled by novel electronic elements like memristors, has shown its potential to sustain the exponential growth of computing demand in the new era of analog data deluge. Here, a platform of a memristive field‐programmable analog array (memFPAA) is experimentally demonstrated with memristive devices serving as a variety of core analog elements and CMOS components as peripheral circuits. The memFPAA is reconfigured to implement a first‐order band pass filter, an audio equalizer, and an acoustic mixed frequency classifier, as application examples. The memFPAA, featured with programmable analog memristors, memristive routing networks, and memristive vector‐matrix multipliers, opens opportunities for fast prototyping analog designs as well as efficient analog applications in signal processing and neuromorphic computing. 

Sneak path current is a fundamental issue and a major roadblock to the wide application of memristor crossbar arrays. Traditional selectors such as transistors compromise the 2D scalability and 3D stack‐ability of the array, while emerging selectors with highly nonlinear current–voltage relations contradict the requirement of a linear current–voltage relation for efficient multiplication by directly using Ohm's law. Herein, the concept of a timing selector is proposed and demonstrated, which addresses the sneak path issue with a voltage‐dependent delay time of its transient switching behavior, while preserving a linear current–voltage relationship for computation. Crossbar arrays with silver‐based diffusive memristors as the timing selectors are built and the operation principle and operational windows are experimentally demonstrated. The timing selector enables large memristor crossbar arrays that can be used to solve large‐dimension real‐world problems in machine intelligence and neuromorphic computing.