Mitigating Imperfections in Mixed-Signal Neuromorphic Circuits
Authors:
Z. Fahimi,
M. R. Mahmoodi,
M. Klachko,
H. Nili,
H. Kim,
D. B. Strukov
Abstract:
The progress in neuromorphic computing is fueled by the development of novel nonvolatile memories capable of storing analog information and implementing neural computation efficiently. However, like most other analog circuits, these devices and circuits are prone to imperfections, such as temperature dependency, noise, tuning error, etc., often leading to considerable performance degradation in ne…
▽ More
The progress in neuromorphic computing is fueled by the development of novel nonvolatile memories capable of storing analog information and implementing neural computation efficiently. However, like most other analog circuits, these devices and circuits are prone to imperfections, such as temperature dependency, noise, tuning error, etc., often leading to considerable performance degradation in neural network implementations. Indeed, imperfections are major obstacles in the path of further progress and ultimate commercialization of these technologies. Hence, a practically viable approach should be developed to deal with these nonidealities and unleash the full potential of nonvolatile memories in neuromorphic systems. Here, for the first time, we report a comprehensive characterization of critical imperfections in two analog-grade memories, namely passively-integrated memristors and redesigned eFlash memories, which both feature long-term retention, high endurance, analog storage, low-power operation, and compact nano-scale footprint. Then, we propose a holistic approach that includes modifications in the training, tuning algorithm, memory state optimization, and circuit design to mitigate these imperfections. Our proposed methodology is corroborated on a hybrid software/experimental framework using two benchmarks: a moderate-size convolutional neural network and ResNet-18 trained on CIFAR-10 and ImageNet datasets, respectively. Our proposed approaches allow 2.5x to 9x improvements in the energy consumption of memory arrays during inference and sub-percent accuracy drop across 25-100 C temperature range. The defect tolerance is improved by >100x, and a sub-percent accuracy drop is demonstrated in deep neural networks built with 64x64 passive memristive crossbars featuring 25% normalized switching threshold variations.
△ Less
Submitted 9 July, 2021;
originally announced July 2021.
Diffusing-Wave Spectroscopy in a Standard Dynamic Light Scattering Setup
Authors:
Zahra Fahimi,
Frank Aangenendt,
Panayiotis Voudouris,
Johan Mattson,
Hans M. Wyss
Abstract:
Diffusing-Wave Spectroscopy (DWS) treats the transport of photons through turbid samples as a diffusion process, thereby making it possible to extract the dynamics of scatterers from measured correlation functions. The analysis of DWS data requires knowledge of the path length distribution of photons traveling through the sample. While for flat sample cells this path length distribution can be rea…
▽ More
Diffusing-Wave Spectroscopy (DWS) treats the transport of photons through turbid samples as a diffusion process, thereby making it possible to extract the dynamics of scatterers from measured correlation functions. The analysis of DWS data requires knowledge of the path length distribution of photons traveling through the sample. While for flat sample cells this path length distribution can be readily expressed in analytical form, no such expression is available for cylindrical sample cells. DWS measurements have therefore typically relied on dedicated setups that use flat sample cells. Here we show how DWS measurements, in particular DWS-based microrheology measurements, can be performed in standard dynamic light scattering setups that use cylindrical sample cells. To do so we perform simple random walk simulations which yield numerical predictions of the path length distribution as a function of both the transport mean free path and the detection angle. This information is used in experiments to extract the mean-square displacement of tracer particles in the material, as well as the resulting frequency-dependent viscoelastic response. An important advantage of our approach is that by measuring at different detection angles, the average photon path length can be varied. Using a single sample cell, this gives access to a wider range of length and time scales than obtained in a conventional DWS setup. Such angle-dependent measurements also offer an important consistency check, as for all detection angles the DWS analysis should yield the same tracer dynamics, even though the respective path length distributions are very different. We validate our approach by performing measurements both on aqueous suspensions of tracer particles and on solid-like gelatin samples, for which we find our DWS-based microrheology data to be in excellent agreement with rheological measurements.
△ Less
Submitted 4 December, 2016; v1 submitted 11 September, 2015;
originally announced September 2015.
