In modern nanometer-scale CMOS technologies, time-zero and time-dependent variability (TDV) effects, the latter coming from aging mechanisms like Bias Temperature Instability (BTI), Hot Carrier Injection (HCI) or Random Telegraph Noise (RTN), have re-emerged as a serious threat affecting the performance of analog and digital integrated circuits. Variability induced by the aging phenomena can lead circuits to a progressive malfunction or failure. In order to understand the effects of the mentioned variability sources, a precise and sound statistical characterization and modeling of these effects should be done. Typically, transistor TDV characterization entails long, and typically prohibitive, testing times, as well as huge amounts of data, which are complex to post-process. In order to face these limitations, this work presents a new method to statistically characterize the emission times and threshold voltage shifts (ΔV_th) related to oxide defects in nanometer CMOS transistors during aging tests. At the same time, the aging testing methodology significantly reduces testing times by parallelizing the stress. The method identifies the Vth drops associated to oxide trap emissions during BTI and HCI aging recovery traces while removing RTN and background noise contributions, to avoid artifacts during data analysis.

Journal Paper - Solid-State Electronics, first online, 2019 ELSEVIER
DOI: 10.1016/j.sse.2019.03.045    ISSN: 0038-1101    » doi

This paper presents an innovative and automated measurement setup for the characterization of variability effects in CMOS transistors using array-based integrated circuits (ICs), through which a better understanding of CMOS reliability could be attained. This setup addresses the issues that come with the need for a trustworthy statistical characterization of these effects: testing a very large number of devices accurately but, also, in a timely manner. The setup consists of software and hardware components that provide a user-friendly interface to perform the statistical characterization of CMOS transistors. Five different electrical tests, comprehending time-zero and time-dependent variability effects, can be carried out. Test preparation is, with the described setup, reduced to a few seconds. Moreover, smart parallelization techniques allow reducing the typically time-consuming aging characterization from months to days or even hours. The scope of this paper thus encompasses the methodology and practice of measurement of CMOS time-dependent variability, as well as the development of appropriate measurement systems and components used in efficiently generating and acquiring the necessary electrical signals.

Journal Paper - IEEE Transactions on Instrumentation and Measurement, first online, 2019 IEEE
DOI: 10.1109/TIM.2019.2906415    ISSN: 0018-9456    » doi

This paper presents a sub-microwatt ac-coupled neural amplifier for the purpose of neural signal sensing. A proposed reconfigurable topology embeds in it filtering capabilities allowing it to select among different frequency bands inside the neural signal spectrum. Power consumption is optimized by designing for bandwidth-specific noise targets that take into account the spectral characteristics of the input signal as well as the noise bandwidths of the noise generators in the circuit itself. An experimentally verified prototype designed in a 180nm CMOS process draws 803nW from a 1V source. The measured input-referred spot-noise at 150Hz is 130nV / Hz while the integrated noise in the 200Hz-5kHz band is 3.6µV rms.

Conference - IEEE Latin American Symposium on Circuits and Systems LASCAS 2019

Artificial Neural Networks (ANNs) show great performance in several data analysis tasks including visual and auditory applications. However, direct implementation of these algorithms without considering the sparsity of data requires high processing power, consume vast amounts of energy and suffer from scalability issues. Inspired by biology, one of the methods which can reduce power consumption and allow scalability in the implementation of neural networks is asynchronous processing and communication by means of action potentials, so-called spikes. In this work, we use the wellknown sigma-delta quantization method and introduce an easy and straightforward solution to convert an Artificial Neural Network to a Spiking Neural Network which can be implemented asynchronously in a neuromorphic platform. Briefly, we used asynchronous spikes to communicate the quantized output activations of the neurons. Despite the fact that our proposed mechanism is simple and applicable to a wide range of different ANNs, it outperforms the state-of-the-art implementations from the accuracy and energy consumption point of view. All source code for this project is available upon request for the academic purpose.

Conference - IEEE International Conference on Artificial Intelligence Circuits and Systems AICAS 2019

This brief presents a piecewise linear approximation of the nonlinear Wilson (NW) neuron model for the realization of an efficient digital circuit implementation. The accuracy of the proposed piecewise Wilson (PW) model is examined by calculating time domain signal shaping errors. Furthermore, bifurcation analyses demonstrate that the approximation follows the same bifurcation pattern as the NW model. As a proof of concept, both models are hardware synthesized and implemented on field programmable gate arrays, demonstrating that the PW model has a range of neuronal behaviors similar to the NW model with considerably higher computational performance and a lower hardware overhead. This approach can be used in hardware-based large scale biological neural network simulations and behavioral studies. The mean normalized root mean square error and maximum absolute error of the PW model are 6.32% and 0.31%, respectively, as compared to the NW model.

Journal Paper - IEEE Transactions on Circuits and Systems II: Express Briefs, vol. 66, no. 1, 2019 IEEE
DOI: 10.1109/TCSII.2018.2852598    ISSN: 1549-7747    » doi