Explore the vast range of titles available.

To circumvent the von Neumann bottleneck, substantial progress has been made towards in-memory computing with synaptic devices. However, compact nanodevices implementing non-linear activation functions are required for efficient full-hardware implementation of deep neural networks. Here, we present an energy-efficient and compact Mott activation neuron based on vanadium dioxide and its successful integration with a conductive bridge random access memory (CBRAM) crossbar array in hardware. The Mott activation neuron implements the rectified linear unit function in the analogue domain. The neuron devices consume substantially less energy and occupy two orders of magnitude smaller area than those of analogue complementary metal–oxide semiconductor implementations. The LeNet-5 network with Mott activation neurons achieves 98.38% accuracy on the MNIST dataset, close to the ideal software accuracy. We perform large-scale image edge detection using the Mott activation neurons integrated with a CBRAM crossbar array. Our findings provide a solution towards large-scale, highly parallel and energy-efficient in-memory computing systems for neural networks. Energy- and area-efficient vanadium-dioxide-based Mott activation neuron devices enable the implementation of activation functions in neural networks.

CMOS-RRAM integration holds great promise for low energy and high throughput neuromorphic computing. However, most RRAM technologies relying on filamentary switching suffer from variations and noise, leading to computational accuracy loss, increased energy consumption, and overhead by expensive program and verify schemes. We developed a filament-free, bulk switching RRAM technology to address these challenges. We systematically engineered a trilayer metal-oxide stack and investigated the switching characteristics of RRAM with varying thicknesses and oxygen vacancy distributions to achieve reliable bulk switching without any filament formation. We demonstrated bulk switching at megaohm regime with high current nonlinearity, up to 100 levels without compliance current. We developed a neuromorphic compute-in-memory platform and showcased edge computing by implementing a spiking neural network for an autonomous navigation/racing task. Our work addresses challenges posed by existing RRAM technologies and paves the way for neuromorphic computing at the edge under strict size, weight, and power constraints. Filamentary RRAM technologies suffer from variations and noise, leading to computational accuracy loss, and increased energy consumption. Park et al. created a trilayer metal-oxide bulk switching RRAM technology without filament formation and showed edge computing for an autonomous navigation task.

Resistive RAM crossbar arrays offer an attractive solution to minimize off-chip data transfer and parallelize on-chip computations for neural networks. Here, we report a hardware/software co-design approach based on low energy subquantum conductive bridging RAM (CBRAM®) devices and a network pruning technique to reduce network level energy consumption. First, we demonstrate low energy subquantum CBRAM devices exhibiting gradual switching characteristics important for implementing weight updates in hardware during unsupervised learning. Then we develop a network pruning algorithm that can be employed during training, different from previous network pruning approaches applied for inference only. Using a 512 kbit subquantum CBRAM array, we experimentally demonstrate high recognition accuracy on the MNIST dataset for digital implementation of unsupervised learning. Our hardware/software co-design approach can pave the way towards resistive memory based neuro-inspired systems that can autonomously learn and process information in power-limited settings. To realize the potential of resistive RAM crossbar arrays as platforms for neuromorphic computing, reduced network-level energy consumption must be achieved. Here, the authors use a hardware/software co-design approach to realize reduced energy consumption during network training for the network.

Inspired from the computational efficiency of the biological brain, spiking neural networks (SNNs) emulate biological neural networks, neural codes, dynamics, and circuitry. SNNs show great potential for the implementation of unsupervised learning using in-memory computing. Here, we report an algorithmic optimization that improves energy efficiency of online learning with SNNs on emerging non-volatile memory (eNVM) devices. We develop a pruning method for SNNs by exploiting the output firing characteristics of neurons. Our pruning method can be applied during network training, which is different from previous approaches in the literature that employ pruning on already-trained networks. This approach prevents unnecessary updates of network parameters during training. This algorithmic optimization can complement the energy efficiency of eNVM technology, which offers a unique in-memory computing platform for the parallelization of neural network operations. Our SNN maintains ~90% classification accuracy on the MNIST dataset with up to ~75% pruning, significantly reducing the number of weight updates. The SNN and pruning scheme developed in this work can pave the way toward applications of eNVM based neuro-inspired systems for energy efficient online learning in low power applications.

Background The nucleotide-binding site–leucine-rich repeat (NBS-LRR) genes are important for plant development and disease resistance. Although genome-wide studies of NBS-encoding genes have been performed in several species, the evolution, structure, expression, and function of these genes remain unknown in radish ( Raphanus sativus L.). A recently released draft R. sativus L. reference genome has facilitated the genome-wide identification and characterization of NBS-encoding genes in radish. Results A total of 225 NBS-encoding genes were identified in the radish genome based on the essential NB-ARC domain through HMM search and Pfam database, with 202 mapped onto nine chromosomes and the remaining 23 localized on different scaffolds. According to a gene structure analysis, we identified 99 NBS-LRR-type genes and 126 partial NBS-encoding genes. Additionally, 80 and 19 genes respectively encoded an N-terminal Toll/interleukin-like domain and a coiled-coil domain. Furthermore, 72% of the 202 NBS-encoding genes were grouped in 48 clusters distributed in 24 crucifer blocks on chromosomes. The U block on chromosomes R02, R04, and R08 had the most NBS-encoding genes (48), followed by the R (24), D (23), E (23), and F (17) blocks. These clusters were mostly homogeneous, containing NBS-encoding genes derived from a recent common ancestor. Tandem (15 events) and segmental (20 events) duplications were revealed in the NBS family. Comparative evolutionary analyses of orthologous genes among Arabidopsis thaliana , Brassica rapa , and Brassica oleracea reflected the importance of the NBS-LRR gene family during evolution. Moreover, examinations of cis-elements identified 70 major elements involved in responses to methyl jasmonate, abscisic acid, auxin, and salicylic acid. According to RNA-seq expression analyses, 75 NBS-encoding genes contributed to the resistance of radish to Fusarium wilt. A quantitative real-time PCR analysis revealed that RsTNL03 ( Rs093020 ) and RsTNL09 ( Rs042580 ) expression positively regulates radish resistance to Fusarium oxysporum , in contrast to the negative regulatory role for RsTNL06 ( Rs053740 ). Conclusions The NBS-encoding gene structures, tandem and segmental duplications, synteny, and expression profiles in radish were elucidated for the first time and compared with those of other Brassicaceae family members ( A. thaliana , B. oleracea , and B. rapa ) to clarify the evolution of the NBS gene family. These results may be useful for functionally characterizing NBS-encoding genes in radish.

Deep learning based on neural networks emerged as a robust solution to various complex problems such as speech recognition and visual recognition. Deep learning relies on a great amount of iterative computation on a huge dataset. As we need to transfer a large amount of data and program between the CPU and the memory unit, the data transfer rate through a bus becomes a limiting factor for computing speed, which is known as Von Neumann bottleneck. Moreover, the data transfer between memory and computation spends a large amount of energy and cause significant delay. To overcome the limitation of Von Neumann bottleneck, neuromorphic computing with emerging nonvolatile memory (eNVM) devices has been proposed to perform iterative calculations in memory without transferring data to a processor. This dissertation presents energy efficient hardware implementation of neuromorphic computing applications using phase change memory (PCM), subquantum conductive bridge random access memory (CBRAM), Ag-based CBRAM, and CuOx-based resistive random access memory (RRAM). Although substantial progress has been made towards in-memory computing with synaptic devices, compact nanodevices implementing non-linear activation functions for efficient full-hardware implementation of deep neural networks is still missing. Since DNNs need to have a very large number of activations to achieve high accuracy, it is critical to develop energy and area efficient implementations of activation functions, which can be integrated on the periphery of the synaptic arrays. In this dissertation, we demonstrate a Mott activation neuron that implements the rectified linear unit function in the analogue domain. The integration of Mott activation neurons with a CBRAM crossbar array is also demonstrated in this dissertation.

The decoy effect is a well-known, intriguing decision-making bias that is often exploited by marketing practitioners to steer consumers towards a desired purchase outcome. It demonstrates that an inclusion of an alternative in the choice set can alter one’s preference among the other choices. Although this decoy effect has been universally observed in the real world and also studied by many economists and psychologists, little is known about how to mitigate the decoy effect and help consumers make informed decisions. In this study, we conducted two experiments: a quantitative experiment with crowdsourcing and a qualitative interview study—first, the crowdsourcing experiment to see if visual interfaces can help alleviate this cognitive bias. Four types of visualizations, one-sided bar chart, two-sided bar charts, scatterplots, and parallel-coordinate plots, were evaluated with four different types of scenarios. The results demonstrated that the two types of bar charts were effective in decreasing the decoy effect. Second, we conducted a semi-structured interview to gain a deeper understanding of the decision-making strategies while making a choice. We believe that the results have an implication on showing how visualizations can have an impact on the decision-making process in our everyday life.

The resistance gene analog (RGA)-based marker strategy is an effective supplement for current marker reservoir of radish disease-resistance breeding. In this study, we identified RGAs based on the conserved nucleotide-binding site (NBS) and S-receptor-like kinase (SRLK) domains. A total of 68 NBS-RGAs and 46 SRLK-RGAs were isolated from two FW-resistant radish inbred lines, B2 and YR31, and one susceptible line, YR15. A BLASTx search revealed that the NBS-RGAs contained six conserved motifs (i.e., P loop, RNBS-A, Kinase-2, RNBS-B, RNBS-C, and GLPL) and the SRLK-RGAs, contained two conserved motifs (i.e., G-type lectin and PAN-AP). A phylogenetic analysis indicated that the NBS-RGAs could be separated into two classes (i.e., toll/interleukin receptor and coiled-coil types), with six subgroups, and the SRLK-RGAs were divided into three subgroups. Moreover, we designed RGA-specific markers from data-mining approach in radish databases. Based on marker analysis, 24 radish inbred lines were clustered into five main groups with a similarity index of 0.44 and showing genetic diversity with resistance variation in those radish inbred lines. The development of RGA-specific primers would be valuable for marker-assisted selection during the breeding of disease-resistant radish cultivars.

A programmable linear resistor with a compact footprint would have profound implications for microelectronics, enabling efficient in-sensor analog signal processing and in-memory computing. Non-volatile memory offers a potential solution but suffers from limitations due to the programming mechanisms that confine switching to nanoscale constrictions or field-sensitive semiconductor junctions, leading to non-linear current-voltage relationships and errors. Here, we introduce a tunable resistor that is programmed into non-volatile, high-precision resistance states spanning nine orders of magnitude, with linear current-voltage characteristics across the entire range -- significantly improving the performance and widening the application space of resistive memory. A key advance is an electrothermal gate that simultaneously spreads heat and electrochemical reactions during programming to enable large, bulk composition modulation. The volumetric modulation can host thousands of linear resistance states with 100x lower conductance errors than other memory. This enables direct processing of analog signals with high fidelity, and we demonstrate variable-gain amplification, division, and multiplication. Integration with CMOS is used to show resilience to electrical and thermal disturb in arrays and to demonstrate retention of analog levels at <1% average loss for more than 2 months across 100 devices. Simulations indicate matrix multiplication efficiency could approach >1,000 TOPS/W.

Resistive memory-based reconfigurable systems constructed by CMOS-RRAM integration hold great promise for low energy and high throughput neuromorphic computing. However, most RRAM technologies relying on filamentary switching suffer from variations and noise leading to computational accuracy loss, increased energy consumption, and overhead by expensive program and verify schemes. Low ON-state resistance of filamentary RRAM devices further increases the energy consumption due to high-current read and write operations, and limits the array size and parallel multiply & accumulate operations. High-forming voltages needed for filamentary RRAM are not compatible with advanced CMOS technology nodes. To address all these challenges, we developed a forming-free and bulk switching RRAM technology based on a trilayer metal-oxide stack. We systematically engineered a trilayer metal-oxide RRAM stack and investigated the switching characteristics of RRAM devices with varying thicknesses and oxygen vacancy distributions across the trilayer to achieve reliable bulk switching without any filament formation. We demonstrated bulk switching operation at megaohm regime with high current nonlinearity and programmed up to 100 levels without compliance current. We developed a neuromorphic compute-in-memory platform based on trilayer bulk RRAM crossbars by combining energy-efficient switched-capacitor voltage sensing circuits with differential encoding of weights to experimentally demonstrate high-accuracy matrix-vector multiplication. We showcased the computational capability of bulk RRAM crossbars by implementing a spiking neural network model for an autonomous navigation/racing task. Our work addresses challenges posed by existing RRAM technologies and paves the way for neuromorphic computing at the edge under strict size, weight, and power constraints.