Search

With the rapid development of digital devices, how to transmit information among different devices with multimedia carrier has drawn much attention from the research community. This paper focuses on the important user scenario ‘`screen-to-camera information transmission’'. Along this direction, image coding based techniques have been shown to be the most popular and effective way in the past decades. However after careful study, we find none of existing methods can satisfy the four important properties simultaneously, i.e., high transparency, high embedding efficiency, strong transmission robustness and high adaptability to device types. It is mainly because these properties are contradictory with each other. So in this paper, we propose an screen-to-camera image code dubbed ‘`TERA’' with Transparency, Efficiency, Robustness and Adaptability, which makes it possible to circumvent the contradiction among the above four properties for the first time. Generally, it adopts the color decomposition principle to ensure the visual quality and the superposition-based scheme to ensure embedding efficiency. And BCH-coding-based information arrangement and a powerful attention-guided information decoding network are further designed to guarantee the robustness and adaptability. Through extensive experiments, the superiority and broad applications of our method are demonstrated.

Xylosandrus compactus and Liparthrum colchicum are two Scolytinae recorded for the first time in Italy in 2011 and 2019, respectively. X. compactus is an “ambrosia beetle” causing damages to several plants typical of the Mediterranean maquis through its digging activity on twigs and branches of the host plants. L. colchicum is a “bark beetle”, monophagous on laurel shrubs and trees. During a survey performed in summer of 2020, in “Tenuta Salviati” located in the Migliarino Natural Park (Pisa, Tuscany), adults of X. compactus and L. colchicum were collected from the same branches and apical twigs of Laurus nobilis hedges. The study aimed at characterizing fungi isolated from the external surface of the two scolytids bodies, to evaluate similarities and even their role in fungal spreading in the environment.

The substitution of nucleotides at specific positions in the genome of a population, known as single-nucleotide polymorphisms (SNPs), has been correlated with a number of important diseases. Complex conditions such as Alzheimer's disease or Crohn's disease are significantly linked to genetics when the impact of multiple SNPs is considered. SNPs often interact in an epistatic manner, where the joint effect of multiple SNPs may not be simply mapped to a linear additive combination of individual effects. Genome-wide association studies considering epistasis are computationally challenging, especially when performing triplet searches is required. Some contemporary computer architectures support fused XOR and population count as the highest throughput operations as part of tensor operations. This article presents a new approach for efficiently repurposing this capability to accelerate 2-way (pairs) and 3-way (triplets) epistasis detection searches. Experimental evaluation targeting the Turing GPU architecture resulted in previously unattainable levels of performance, with the proposal being able to evaluate up to 108.1 and 54.5 tera unique sets of SNPs per second, scaled to the sample size, in 2-way and 3-way searches, respectively.

This study aims to determine the most dominant factor affecting the purchasing decision of Indocafe powder coffee at PT. Sari Tanjung Tera Cipta in Bengkulu. Samples in this study were 37 outlets that purchased Indocafe powder coffee products at PT. Sari Tanjung Tera Cipta in Bengkulu City, especially the Bengkulu route. The data collection method used a questionnaire and the analytical method used was rating scale analysis. The results of this study indicate the average value of respondents' responses to the factors that affect purchasing decisions at PT. Sari Tanjung Tera Cipta in Bengkulu City is 1,474 with the assessment criteria agree. This means that respondents have an agreed assessment of the factors that affect purchasing decisions at PT. Sari Tanjung Tera Cipta in Bengkulu City is affected by cultural, social, personal and psychological factors. Personal factors are the highest factors influencing purchasing decisions online at PT. Sari Tanjung Tera Cipta in Bengkulu City with a score of 1,573. This illustrates that customers shop at PT. Sari Tanjung Tera Cipta in Bengkulu City because because they choose to shop on the basis of their own wishes, PT. Sari Tanjung Tera Cipta in Bengkulu City provides maximum service therefore customers feel satisfied. The lowest factor is the cultural factor with a total score of 1.342 with the assessment criteria agreed. This illustrates that cultural factors have a low affect on purchasing decisions at PT. Sari Tanjung Tera Cipta in Bengkulu City because shopping is a desire not because it is affected by the environment, culture or other people.

Physical fitness or cardiovascular fitness (VO2Max) is a determining factor for the health status of the pilgrims. Cardiovascular fitness can be improved in several ways, including exercise and nutrition management.This study aims to determine the effectiveness of exercise and nutritional management on the heart and lung fitness of the pilgrims. Quasi-experimentalmethodusing pretest-posttest non-equivalent control group design and using the t-test on 34 prospective pilgrims showed. This study usedtheRockpot Instrument.Result:Therewas a significant effect between exercise and nutritional management on cardiovascular fitness. (P-Value: 0.05 and 0.00,α: 0.05).Conclution:The variablethat has themostinfluenceoncardiovascularfitnessisexercisemodifiedwithnutritionalmanagement. Gymnastics exercises and counseling on nutrition using video lessons are very effective in improving heart-lung fitness for aspiring pilgrims.

Terahertz Metamaterials are a type of Metamaterials, capable of interacting at terahertz (THz) frequencies. The bibliometric analysis had been conducted to understand the active authors, organizations, journals, and countries involved in the research domain of “Terahertz Metamaterials”. All published articles related to “Terahertz Metamaterials” from “Scopus”, were analyzed using the VOS viewer to develop analysis tables and visualization maps.This article had set the objective to consolidate the scientific literature regarding the “Terahertz Metamaterials”and also to find out the trends related to the same.The most active journals in this research domain were Optics Express and Nature Nanotechnology. The most active countries were China, Greece, and the United States of America. The leading organizations were the Ministry of Education of China and Tianjin University of China.The most active authorswere Zhang X., Zhang W., and Averitt R.D.

Objectives The present study was conducted to evaluate the usefulness of capillary electrophoresis (CE), cation-exchange high-performance liquid chromatography (HPLC), and matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry (MS) for the screening of hemoglobin (Hb) variants prevalent in southern China. Methods A total of 102 types of Hb variants in 1,083 variant carriers were identified over a 5-year period. These variants were analyzed by a CE method (Capillarys 3 TERA), a cation-exchange HPLC analyzer (Variant II Turbo 2.0), and a MALDI-TOF MS system (QuanTOF). Results The presence of 85 (83.3%, 85/102), 84 (82.4%, 84/102), and 62 (60.8%, 62/102) Hb variants was detected by Capillarys 3 TERA, Variant II Turbo 2.0, and QuanTOF, respectively. Of the three methods, only Capillarys 3 TERA recognized all 10 of the most frequent Hb variants in southern China. There were six, two, and three Hb variants that can only be detected by Capillarys 3 TERA, Variant II Turbo 2.0, and QuanTOF, respectively. The detection limit of mass difference for QuanTOF was approximately 11 to 20 Da. Conclusions MALDI-TOF MS is suitable for use as an auxiliary method rather than a stand-alone method for the screening of Hb variants prevalent in southern China.

In this work, a heterogeneous integration strategy of 3D NAND based compute-in-memory (CIM) architecture is proposed for large-scale deep neural networks (DNNs). While most of the reported CIM architectures today have focused on the image classification models with MB-level parameters, we aim at huge language translation models with GB-scale parameters. Our 3D NAND CIM architecture design exploits two fabrication techniques, wafer bonding scheme and CMOS under array (CUA), to integrate CMOS circuits, 3D NAND cells, and high voltage (HV) transistors at different tiers without thermal budget issue during the fabrication process. The bonding pads between two wafers are designed to transfer the input and output vectors while ensuring $\sim 1~\mu \text{m}$ pitch that is feasible by hybrid bonding. The chip size of the 512 Gb 128-layer 3D NAND CIM architecture is estimated to be 166 mm2 with 7 nm FinFET logic transistors. Using the physical and electrical parameters of standard 3D NAND cells, the 1.15-19.01 tera operations per second per watt (TOPS/W) of energy efficiency is achieved.

Computationally intensive Inference tasks of Deep neural networks have enforced revolution of new accelerator architecture to reduce power consumption as well as latency. The key figure of merit in hardware inference accelerators is the number of multiply-and-accumulation operations per watt (MACs/W), where, the state-of-the-arts MACs/W remains several hundreds Giga-MACs/W. We propose a Tera-MACS/W neural hardware inference Accelerator (TMA) with 8-bit activations and scalable integer weights less than 1-byte. The architectures main feature is configurable neural processing element for matrix-vector operations. The proposed neural processing element has Multiplier-less Massive Parallel Processor to work without any multiplications, which makes it attractive for energy efficient high-performance neural network applications. We benchmark our systems latency, power, and performance using Alexnet trained on ImageNet. Finally, we compared our accelerators throughput and power consumption to the prior works. The proposed accelerator outperforms the state of the art in terms of energy and area achieving 2.3 TMACS/W@1.0 V, 65 nm CMOS technology.

Performance of biomedical analog circuits vitiates due to non-linear V-R characteristics of Pseudo-resistor (PR) structure, their excess dependency over Process Voltage Temperature (PVT) and common mode variations. In this paper, Incremental Resistance (IR) expressions, V-R curves and statistical results for wider PVT variations for diverse categories of non-tunable and tunable PR structures have been presented. The results obtained using 0.18 μm standard CMOS process with BSIM3V3 device models show that Dynamic Threshold Metal Oxide Semiconductor (DTMOS) technique involving symmetrical biasing for bulk and gate voltages for MOS devices is capable to emulate high nominal value (RYX, N) of IR of the order of 1.95 TΩ over a wider voltage swing of −1 V to 1 V with an area consumption of 60 μm2. Statistical results obtained after performing Monte Carlo Simulations (MCS) with 100 runs for the verification of robustness against PVT variations for this structure shows the value of mean (μ), standard deviation (σ) and Coefficient of variance (CV = σ/μ) as 1.918 TΩ, 100.6 GΩ and 0.052 respectively. Results for a variety of other PR structures reported in this paper can be useful for their possible deployment in circuit applications like neural signal recording and low current sensing etc.

Connected filters and multi-scale tools are region-based operators acting on the connected components of an image. Component trees are image representations to efficiently perform these operations as they represent the inclusion relationship of the connected components hierarchically. This paper presents disccofan (DIStributed Connected COmponent Filtering and ANalysis), a new method that extends the previous 2-D implementation of the Distributed Component Forests (DCFs) to handle 3-D processing and higher dynamic range data sets. disccofan combines shared and distributed memory techniques to efficiently compute component trees, user-defined attributes filters, and multi-scale analysis. Compared to similar methods, disccofan is faster and scales better on low and moderate dynamic range images, and is the only method with a speed-up larger than 1 on a realistic, astronomical floating-point data set. It achieves a speed-up of 11.20 using 48 processes to compute the DCF of a 162 Gigapixels, single-precision floating-point 3-D data set, while reducing the memory used by a factor of 22. This approach is suitable to perform attribute filtering and multi-scale analysis on very large 2-D and 3-D data sets, up to single-precision floating-point value.

Complementary metal oxide semiconductor (CMOS) devices are expected to face new challenges such as exponential current leakage, DIBL, hot carrier effects, and etc. at nano scale. Hence the CMOS technology is being supplanted by the nanotechnologies. Quantum-dot Cellular Automata (QCA) is key technology at nano scale which operates at tera hertz of speed. Comparing with the traditional CMOS technology, the QCA technology has low power consumption and high density. This technology also has a unique methodology such as “processing in wire” and “memory-in-motion”. This work confers an area proficient, high speed full adder (FA) design with efficient clocking. The proposed full adder design comprises of 26 Quantum cells with delay of two clock phases, area occupancy of 0.03 µm2. This paper utilizes the unique characteristics of QCA and the proposed Wallace tree multiplier is implemented with the mentioned FA and is designed by using QCA Designer 2.0.3 tool.

In this work we propose an asynchronous, GPU-based implementation of the widely-used stochastic coordinate descent algorithm for convex optimization. We define the class of problems that can be solved using this method, and show that it includes many popular machine learning applications. For three such applications, we demonstrate at least a 10 × training speed-up relative to a state-of-the-art implementation that uses all available resources of a modern CPU. In order to train on very large datasets that do not fit inside the memory of a single GPU, we then consider techniques for distributed learning. We show that while such techniques do not necessarily allow one to achieve further speed-up, they do allow one to train on datasets that would otherwise not fit into memory. We thus propose a distributed learning system that uses the synchronous CoCoA framework to distribute the global optimization across GPUs, and our novel asynchronous algorithm to solve the corresponding local optimizations within each GPU. We benchmark such a system using a 200 GB dataset that consists of 1 billion training examples. We show by scaling out across 16 GPUs, we can train an SVM model to a high degree of accuracy in around 1 min: a 15 × speed-up in training time compared to a state-of-the-art CPU-based implementation that uses 640 threads distributed across 8 CPUs.

Objective Previous research has found increasing evidence for difficulties in emotion recognition ability (ERA) and social cognition in anorexia nervosa (AN), and recent models consider these factors to contribute to the development and maintenance of the disorder. However, there is a lack of experimental studies testing this hypothesis. Therefore, the present proof-of-concept study examined whether ERA can be improved by a single session of a computerized training in AN, and whether this has short-term effects on eating disorder symptoms. Method Forty inpatients (22.20 ± 7.15 years) with AN were randomly assigned to receive a single session of computerized training of ERA (TERA) or a sham training (training the recognition of different types of clouds). ERA, self-reported eating disorder symptoms, and body mass index (BMI) were assessed within 3 days before and after training. Results After training, both groups showed improved ERA, reduced self-reported eating disorder symptoms, and an increased BMI. As compared to patients in the control group, patients who received TERA showed greater improvements in ERA and self-reported eating disorder symptoms. Discussion ERA can be effectively trained in patients with AN. Moreover, short-term improvements in self-reported eating disorder symptoms provide tentative support for the hypothesis that difficulties in ERA contribute to the maintenance of AN, and that specific trainings of ERA hold promise as an additional component in AN treatment. Future studies are needed to replicate these findings in larger samples, and to investigate long-term effects and transfer into real-world settings.

Cardinality estimation is the task of determining the number of distinct elements (or the cardinality) in a data stream, under a stringent constraint that the input data stream can be scanned by just one single pass. This is a fundamental problem with many practical applications, such as traffic monitoring of high-speed networks and query optimization of Internet-scale database. To solve the problem, we propose an algorithm named HLL-TailCut, which implements the estimation standard error $1.0 / \sqrt {m}$ using the memory units of four or three bits each, whose cost is much smaller than the five-bit memory units used by HyperLogLog, the best previously known cardinality estimator. This makes it possible to reduce the memory cost of HyperLogLog by 20%~45%. For example, when the target estimation error is 1.1%, state-of-the-art HyperLogLog needs 5.6 kilobytes memory. By contrast, our new algorithm only needs 3 kilobytes memory consumption for attaining the same accuracy. Additionally, our algorithm is able to support the estimation of very large stream cardinalities, even on the Tera and Peta scale.

Spodoptera litura, an important polyphagous pest, causes serious threat to a large number of economically important crops all over the world. In the present study, the quantitative analysis of aminoacids and carbohydrates in haemolymph of the last instar larvae of Spodoptera litura fed with five different host plant leaves was carried out. A comparative study of the biochemical components in the selected host plant leaves was also carried out. The data obtained were statistically analysed using one-way ANOVA (P

Although External Beam Radiation Therapy (EBRT) is essential tool for the radiation therapy of cervical cancer; only one cancer institute in Tanzania performs 3-Dimensional Conformal Radiation Therapy (3DCRT) Computed Tomography (CT)-based planning. To identify benefits and advantages of 3D-CRT over 2D- conventional radiation therapy (2D-CRT), dosimetric parameters for tumor targets and organs at risk (OARs) were compared between these modalities for 23 cervical cancer patients. 11 cervical cancer patients were CT scanned after proper positioning and immobilization and transferred to Eclipse Treatment Planning System (TPS) for dose planning. The remaining 12 curative intent patients were planned using 2D-CRT system and treatment times were calculated for each patient. From the CT based planning, the minimum dose (D min), maximum dose (D max) and mean dose (D mean) to Planning Target Volume (PTV) and organs at risk (OAR), were compared for each plan. On average, the optimized maximum doses for bladder, rectum, femoral heads, PTV and Gross Tumor Volume (GTV) were 46.56 Gy, 42.65 Gy, 28.76 Gy, 48.56 Gy and 48.53 Gy. For 2D-concentional planning, the dose rate was 75.75 cGy/min and the average treatment time was 1.6075 minutes. This study confirms that 3D CT-based planning is a good choice in the treatment protocol for carcinoma cervix as it delivered a highly homogeneous and conformal plan with superior dose coverage to PTV and better OARs sparing.

ilustraciones, fotografias La investigación describe cómo un programa piloto de musicoterapia favorece el desarrollo de acciones comunitarias en un grupo de mujeres voluntarias de la fundación She Is en El Espinal-Tolima. Este trabajo presenta un enfoque de investigación cualitativa con un diseño basado en la Investigación Acción Participativa y se siguieron cuatro fases de implementación del programa piloto en las que se aplicaron herramientas para la recolección de datos como la entrevista abierta, grupos focales, diario de campo, línea de tiempo y una lista de chequeo para el seguimiento de las acciones comunitarias con el equipo de voluntariado. El desarrollo del programa piloto de musicoterapia se realizó de manera virtual durante 14 encuentros de 2 horas cada uno, en un periodo equivalente a 14 semanas. Durante este periodo, el equipo de voluntariado propuso un total de 19 acciones comunitarias. De esas acciones, 11 fueron completadas, 1 quedó en proceso, 4 quedaron pendientes a modo de proyección en el corto, mediano y largo plazo y 3 acciones fueron descartadas. Gracias a las acciones completadas surgieron dos producciones musicales significativas para el equipo por su contribución a la proyección y visibilización de sus propósitos ante una comunidad más amplia. De la investigación se concluye que la contribución de musicoterapia al desarrollo de acciones comunitarias en esta población ocurre a través de: a) la delimitación de objetivos en comunidad, b) el seguimiento de acciones comunitarias, c) el uso de los cuatro métodos de la musicoterapia, d) la producción de las composiciones elaboradas por el equipo de voluntariado, e) un proceso como contexto para el encuentro recurrente del equipo de voluntariado, f) el acompañamiento de una musicoterapeuta que comprende el contexto del territorio, g) el abordaje del proceso de musicoterapia desde un enfoque comunitario, h) la apertura para recibir a nuevas personas que se integran al equipo e i) una metodología de investigación coherente con el objetivo. The research describes how a pilot music therapy program favors the development of community actions in a group of women volunteers of the She Is Foundation at El Espinal-Tolima. This work presents a qualitative research approach with a design based on Participatory Action Research and followed four phases of implementation of the pilot program in which tools were applied for data collection such as open interviews, focus groups, field diary, timeline and a checklist for monitoring community actions with the volunteer team. The development of the pilot music therapy program was carried out virtually during 14 2- hour meetings over a 13-week period. During this period of time, the volunteer team proposed a total of 19 community actions. Out of those actions, 11 were completed, 1 is still in process, 4 are pending and projected for the short, medium and long term, and 3 actions were discarded. Two music productions that were meaningful for the team emerged from the completed actions. These music productions contributed to the projection and visibility of the team’s purposes before a wider community. From the research it is concluded that the contribution of music therapy to the development of community actions in this population occurs through: a) the delimitation of objectives in community; b) the monitoring of community actions; c) the use of the four music therapy methods by the music therapist; d) the production of the compositions elaborated by the volunteer team; e) a process as a context for the recurrent meeting of the volunteer team; f) the accompaniment of a music therapist who understands the territory’s context; g) the engaging in the music therapy process through a community approach, h) the openness to receive new people who join the team, and i) a research methodology consistent with the objective. Maestría Magíster en Musicoterapia La presente investigación se plantea desde un enfoque cualitativo con un diseño basado en la Investigación Acción Participativa (IAP). Musicoterapia Comunitaria

Diabetes is a chronic metabolic disorder characterized by hyperglycemia and associated with many health complications due to the long-term damage and dysfunction of various organs. A consequential complication of diabetes in men is reproductive dysfunction, reduced fertility, and poor reproductive outcomes. However, the molecular mechanisms responsible for diabetic environment-induced sperm damage and overall decreased reproductive outcomes are not fully established. We evaluated the effects of type 2 diabetes exposure on the reproductive system and the reproductive outcomes of males and their male offspring, using a mouse model. We demonstrate that paternal exposure to type 2 diabetes mediates intergenerational and transgenerational effects on the reproductive health of the offspring, especially on sperm quality, and on metabolic characteristics. Given the transgenerational impairment of reproductive and metabolic parameters through two generations, these changes likely take the form of inherited epigenetic marks through the germline. Our results emphasize the importance of improving metabolic health not only in women of reproductive age, but also in potential fathers, in order to reduce the negative impacts of diabetes on subsequent generations.

The stochastic nature of many Machine Learning (ML) algorithms makes testing of ML tools and libraries challenging. ML algorithms allow a developer to control their accuracy and run-time through a set of hyper-parameters, which are typically manually selected in tests. This choice is often too conservative and leads to slow test executions, thereby increasing the cost of regression testing. We propose TERA, the first automated technique for reducing the cost of regression testing in Machine Learning tools and libraries(jointly referred to as projects) without making the tests more flaky. TERA solves the problem of exploring the trade-off space between execution time of the test and its flakiness as an instance of Stochastic Optimization over the space of algorithm hyper-parameters. TERA presents how to leverage statistical convergence-testing techniques to estimate the level of flakiness of the test for a specific choice of hyper-parameters during optimization. We evaluate TERA on a corpus of 160 tests selected from 15 popular machine learning projects. Overall, TERA obtains a geo-mean speedup of 2.23x over the original tests, for the minimum passing probability threshold of 99%. We also show that the new tests did not reduce fault detection ability through a mutation study and a study on a set of 12 historical build failures in studied projects.

We report a new approach to ONNs based on integrated Kerr micro-combs that is programmable, highly scalable and capable of reaching ultra-high speeds. We demonstrate a single neuron perceptron at 11.9 Giga-OPS at 8 bits per OP, or 95.2 Gbps. We then demonstrate a convolutional accelerator operating beyond 11 TeraOPs/s (TOPs). We test the perceptron on handwritten-digit recognition and cancer-cell detection — achieving over 90% and 85% accuracy, respectively.

There has been demonstrated a sharp increase of chemical pressing on the environment and human health, detection of hundreds of chemical compounds in different environmental objects, most of such chemicals have no hygienic standards. There are presented main disadvantages oKikuworks on the risks assessment of the impact ofpolluted environment on human health. There are indicated priority directions of the improvement of the analysis methodology and risk management, based on modern international achievements, as well as evaluation of detriments to the environment and human health with taking into account world systems as follows: AirQ (WHO), IEHIA and APHEIS (EU), FERET and EPA (USA), EAHEAP and COMEAP (GreatBritain), ECOSENSE (Germany), AirPack (EU, France), AQVM (Canada), and also domestic of TERA 2,5 (module EpidRisk). The integral evaluation of the scientific disciplines “Human Ecology”, “EnvironmentalHealth” and “EnvironmentalMedicine” is given. Comparative conceptual considertion of the terms “Environment”, “Habitat” and their international application is given.

Hypertension is a cardiovascular disease globally. Hypertension is remains silent killer, the clinical strategy to focusing on new and improved treatments is exercise. Tera gymnastics is a physical and mental exercise, combining the movement of body parts with breathing techniques and rhythms through the concentration of thought that is carried out regularly, harmoniously, correctly and continuously, Physical activity can reduce high blood pressure. Some study showed Regular physical activity is an effective intervention with respect to these factor, decreasing mortality rate for cardiovascular disease and all cause of disease in hypertensive patient. Gymnastic Tera exercise can help to control metabolic variable related to hypertension. The study was use quasy experimental design with one group pretest-posttest. The study was conducted in Puskesmas Pasundan Garut. Sampling in this study is subjects that is criteria patients has hipertension with 8 weeks treatment so will take methode of concecutive sampling with 15 respondents for intervention group and 15 for control group. The sample were age 30-55 years. The Intervention was gymnastic tera exercise. The Blood pressure function was evaluated before and after the training period. The data were analyzed by using t-test paired. The result showed a significant difference before and after tera gymnastic exercise in patient with hypertension I (p=000.0). Result of systolic blood presure when in mean SD pretest is 146.00 higher than post test is 136.00 in intervention group. The tera gymnastic exercise conditioning program achieved effect in this population. The reduction of blood pressure after exercise is of great clinical relevance. The increased riskof physical inactivity in controlling hypertension in our study suggest that general practitioners must be in the habit of prescribing practice of physical exercise. Physical activity has been shown to have beneficial effect on blood pressure. patients are followed up regularly to confirm that they are adhering to the management plan and the blood pressure targets.

This paper describes some achievements of the low-level radio frequency (LLRF) tests performed to evaluate the superconducting cryomodules in preparation of their installation in the European X-ray Free Electron Laser (XFEL) accelerator. The software developed to characterize the cryomodules tested at the Accelerator Module Test Facility (AMTF) will be presented. The purpose of these tests is to evaluate some of the key parameters of the eight Tera Electronvolt Superconducting Linear Accelerator cavities, housed in every cryomodule. These parameters include: cavity quench gradients, cavity fundamental resonant modes, performance of the slow and fast frequency tuners, and characteristics of the input power coupler. These results impact the final decision on the cryomodule acceptance and its installation inside the linear accelerator (linac). In this approach, the MicroTCA.4-based LLRF system is used to control and assess the cryomodule performance. Middle layer servers are developed to verify the tests initial conditions, carry the test, and log the results in a dedicated database. The tests results provide a basis for the cryomodule validation, but also serve as a reference useful during the commissioning and operating phase of the accelerator.

In the present paper rectangular patch array antenna on circular plane for Tera Hertz applications has been analyzed in terms of directivity and scan angle with different radius of circular plane. All the analysis are carried at 1THz frequency for 3 × 3, 5 × 5, 7 × 7 and 9 × 9 antenna array structure. It is found that directivities are maximum at broadside angle direction, and drop sharply with scan angle if number of elements is less. With increase in number of array elements directivity increases, and it is found that for maximum directivity, the radius of circular arc should also be increased. The analysis of directivities and Scanning angles are obtained using MATLAB.

A rigorous and accurate method is presented for the efficient computation of the directive surface plasmon polaritons excited in conductive hyperbolic metasurfaces residing in a homogeneous medium by a vertical Hertzian dipole operating in the terahertz range. Sample numerical results are included that validate the formulation and the associated computer code.

Convolutional neural networks (CNNs), inspired by biological visual cortex systems, are a powerful category of artificial neural networks that can extract the hierarchical features of raw data to greatly reduce the network parametric complexity and enhance the predicting accuracy. They are of significant interest for machine learning tasks such as computer vision, speech recognition, playing board games and medical diagnosis [1-7]. Optical neural networks offer the promise of dramatically accelerating computing speed to overcome the inherent bandwidth bottleneck of electronics. Here, we demonstrate a universal optical vector convolutional accelerator operating beyond 10 Tera-FLOPS (floating point operations per second), generating convolutions of images of 250,000 pixels with 8-bit resolution for 10 kernels simultaneously — enough for facial image recognition. We then use the same hardware to sequentially form a deep optical CNN with ten output neurons, achieving successful recognition of full 10 digits with 900 pixel handwritten digit images with 88% accuracy. Our results are based on simultaneously interleaving temporal, wavelength and spatial dimensions enabled by an integrated microcomb source. This approach is scalable and trainable to much more complex networks for demanding applications such as unmanned vehicle and real-time video recognition.

Convolutional neural networks (CNNs), inspired by biological visual cortex systems, are a powerful category of artificial neural networks that can extract the hierarchical features of raw data to greatly reduce the network parametric complexity and enhance the predicting accuracy. They are of significant interest for machine learning tasks such as computer vision, speech recognition, playing board games and medical diagnosis [1-7]. Optical neural networks offer the promise of dramatically accelerating computing speed to overcome the inherent bandwidth bottleneck of electronics. Here, we demonstrate a universal optical vector convolutional accelerator operating beyond 10 Tera-OPS (TOPS - operations per second), generating convolutions of images of 250,000 pixels with 8-bit resolution for 10 kernels simultaneously — enough for facial image recognition. We then use the same hardware to sequentially form a deep optical CNN with ten output neurons, achieving successful recognition of full 10 digits with 900 pixel handwritten digit images with 88% accuracy. Our results are based on simultaneously interleaving temporal, wavelength and spatial dimensions enabled by an integrated microcomb source. This approach is scalable and trainable to much more complex networks for demanding applications such as unmanned vehicle and real-time video recognition.

The paper describes design and analysis of Tera Hertz Optical Asymmetric Demultiplexer (TOAD) switch using quantum dot semiconductor optical amplifier (QDSOA) and all- optical digital to analog converter (DAC). The QDSOA based TOAD or QDSTOAD and the DAC performance optimization for control power variations shows that 1.5 mW pump or control power gives the optimized performance as far as the extinction ratio is concerned for the TOAD. At the same power absolute error (e) and dynamic range for DAC also shows maximum. The present paper investigates relative output, absolute error and dynamic range for the performance analysis of the DAC. An absolute error as low as 0.25% and dynamic range (11.765 dB) very close to the ideal value (11.761 dB) ensures effectiveness of these devices. These devices show immunity to the amplified spontaneous emission noise.

This work presents the design of a constant, programmable and extremely high (Tera ohm) MOS pseudo-resistor over a wide dynamic range of ± 0.6 V based on MOS transistors operating in the weak inversion region. The major challenge faced by the circuit designers in designing high resistance pseudo-resistor is the linearity problem and it is resolved in this work. The proposed pseudo-resistor is utilized as a feedback resistor in the second stage of the two-stages bio-medical amplifiers in the portable bio-detection system. By controlling the resistance of this pseudo-resistor, the lower cut-off frequency of the amplifier is automatically adjusted and thus allowing the detection of the small frequency range. Simulations in LT-spice using 130 nm CMOS technology under ± 0.6 V supply voltage are performed to validate the realization of the extremely high resistance and its application.

Convolutional neural networks (CNNs), inspired by biological visual cortex systems, are a powerful category of artificial neural networks that can extract the hierarchical features of raw data to greatly reduce the network parametric complexity and enhance the predicting accuracy. They are of significant interest for machine learning tasks such as computer vision, speech recognition, playing board games and medical diagnosis [1-7]. Optical neural networks offer the promise of dramatically accelerating computing speed to overcome the inherent bandwidth bottleneck of electronics. Here, we demonstrate a universal optical vector convolutional accelerator operating beyond 10 Tera-OPS (TOPS - operations per second), generating convolutions of images of 250,000 pixels with 8-bit resolution for 10 kernels simultaneously — enough for facial image recognition. We then use the same hardware to sequentially form a deep optical CNN with ten output neurons, achieving successful recognition of full 10 digits with 900 pixel handwritten digit images with 88% accuracy. Our results are based on simultaneously interleaving temporal, wavelength and spatial dimensions enabled by an integrated microcomb source. This approach is scalable and trainable to much more complex networks for demanding applications such as unmanned vehicle and real-time video recognition.

Convolutional neural networks (CNNs), inspired by biological visual cortex systems, are a powerful category of artificial neural networks that can extract the hierarchical features of raw data to greatly reduce the network parametric complexity and enhance the predicting accuracy. They are of significant interest for machine learning tasks such as computer vision, speech recognition, playing board games and medical diagnosis [1-7]. Optical neural networks offer the promise of dramatically accelerating computing speed to overcome the inherent bandwidth bottleneck of electronics. Here, we demonstrate a universal optical vector convolutional accelerator operating beyond 10 Tera-OPS (TOPS - operations per second), generating convolutions of images of 250,000 pixels with 8-bit resolution for 10 kernels simultaneously — enough for facial image recognition. We then use the same hardware to sequentially form a deep optical CNN with ten output neurons, achieving successful recognition of full 10 digits with 900 pixel handwritten digit images with 88% accuracy. Our results are based on simultaneously interleaving temporal, wavelength and spatial dimensions enabled by an integrated microcomb source. This approach is scalable and trainable to much more complex networks for demanding applications such as unmanned vehicle and real-time video recognition.

As Machine Learning (ML) becomes pervasive in the era of artificial intelligence, ML specific tools and frameworks are required for architectural research. This paper introduces NeuroMeter, an integrated power, area, and timing modeling framework for ML accelerators. NeuroMeter models the detailed architecture of ML accelerators and generates a fast and accurate estimation on power, area, and chip timing. Meanwhile, it also enables the runtime analysis of system-level performance and efficiency when the runtime activity factors are provided. NeuroMeter’s micro-architecture model includes fundamental components of ML accelerators, including systolic array based tensor units (TU), reduction trees (RT), and 1D vector units (VU). NeuroMeter has accurate modeling results, with the average power and area estimation errors below 10% and 17% respectively when validated against TPU-v1, TPU-v2, and Eyeriss.Leveraging the NeuroMeter’s new capabilities on architecting manycore ML accelerators, this paper presents the first in-depth study on the design space and tradeoffs of “Brawny and Wimpy” inference accelerators in datacenter scenarios with the insights that are otherwise difficult to discover without NeuroMeter. Our study shows that brawny designs with 64x64 systolic arrays are the most performant and efficient for inference tasks in the 28nm datacenter architectural space with a 500mm2 die area budget. Our study also reveals important tradeoffs between performance and efficiency. For datacenter accelerators with low batch inference, a small $(\sim 16$%) sacrifice of system performance (in achieved Tera OPerations per Second, aka TOPS) can lead to more than a 2x efficiency improvement (in achieved TOPS/TCO). To showcase NeuroMeter’s capability to model a wide range of diverse ML accelerator architectures, we also conduct a followon mini-case study on implications of sparsity on different ML accelerators, demonstrating wimpier accelerator architectures benefit more readily from sparsity processing despite their lower achievable raw energy efficiency.

Tera Hertz Optical asymmetric Demultiplexer or TOAD is an important optical switch in optical computational and communication applications. It is used to design different types of logic gates and logic circuits. In this chapter, we have utilized it to design a frequency-encoded optical XOR gate, half adder, and half subtractor using TOAD-based switch for the first time. The input/output bit patterns along with ON/OFF ratio is also investigated.

We propose a co-design approach for compute-in-memory inference for deep neural networks (DNN). We use multiplication-free function approximators based on $\ell _{1}$ norm along with a co-adapted processing array and compute flow. Using the approach, we overcame many deficiencies in the current art of in-SRAM DNN processing such as the need for digital-to-analog converters (DACs) at each operating SRAM row/column, the need for high precision analog-to-digital converters (ADCs), limited support for multi-bit precision weights, and limited vector-scale parallelism. Our co-adapted implementation seamlessly extends to multi-bit precision weights, it doesn’t require DACs, and it easily extends to higher vector-scale parallelism. We also propose an SRAM-immersed successive approximation ADC (SA-ADC), where we exploit the parasitic capacitance of bit lines of SRAM array as a capacitive DAC. Since the dominant area overhead in SA-ADC comes due to its capacitive DAC, by exploiting the intrinsic parasitic of SRAM array, our approach allows low area implementation of within-SRAM SA-ADC. Our $8\times 62$ SRAM macro, which requires a 5-bit ADC, achieves ~105 tera operations per second per Watt (TOPS/W) with 8-bit input/weight processing at 45 nm CMOS. Our $8\times 30$ SRAM macro, which requires a 4-bit ADC, achieves ~84 TOPS/W. SRAM macros that require lower ADC precision are more tolerant of process variability, however, have lower TOPS/W as well. We evaluated the accuracy and performance of our proposed network for MNIST, CIFAR10, and CIFAR100 datasets. We chose a network configuration which adaptively mixes multiplication-free and regular operators. The network configurations utilize the multiplication-free operator for more than 85% operations from the total. The selected configurations are 98.6% accurate for MNIST, 90.2% for CIFAR10, and 66.9% for CIFAR100. Since most of the operations in the considered configurations are based on proposed SRAM macros, our compute-in-memory’s efficiency benefits broadly translate to the system-level.

We report a new approach to ONNs based on integrated Kerr micro-combs that is programmable, highly scalable and capable of reaching ultra-high speeds. We demonstrate a single neuron perceptron at 11.9 Giga-OPS at 8 bits per OP, or 95.2 Gbps. We then demonstrate a convolutional accelerator operating beyond 11 TeraOPs/s. We test the perceptron on handwritten-digit recognition and cancer-cell detection — achieving over 90% and 85% accuracy, respectively.