Accelerating On-Chip Training with Ferroelectric-Based Hybrid Precision Synapse

In this article, we propose a hardware accelerator design using ferroelectric transistor (FeFET)-based hybrid precision synapse (HPS) for deep neural network (DNN) on-chip training. The drain erase scheme for FeFET programming is incorporated for both FeFET HPS design and FeFET buffer design. By using drain erase, high-density FeFET buffers can be integrated onchip to store the intermediate input-output activations and gradients, which reduces the energy consuming off-chip DRAM access. Architectural evaluation results show that the energy efficiency could be improved by 1.2× ∼ 2.1×, 3.9× ∼ 6.0× compared to the other HPS-based designs and emerging non-volatile memory baselines, respectively. The chip area is reduced by 19% ∼ 36% compared with designs using SRAM on-chip buffer even though the capacity of FeFET buffer is increased. Besides, by utilizing drain erase scheme for FeFET programming, the chip area is reduced by 11% ∼ 28.5% compared with the designs using body erase scheme.

Environmental Performance Assessment of Energy-Consuming Sectors Through Novel Data Envelopment Analysis

This study evaluates the role of information in the environmental performance index (EPI) in different energy-consuming sectors in Pakistan through a novel slack-based data envelopment analysis (DEA). The index combines energy consumption as the primary input and gross domestic product (GDP) as the desirable output and CO2 emissions as the undesirable output. Yale’s EPI measures the efficiency of the sectoral level environmental performance of primary energy consumption in the country. Performance analysis was conducted from 2009 to 2018. The sectors were assigned scores between one and zero, with zero indicating maximum decision-making unit (DMU) inefficiency and one indicating maximum DMU efficiency. Despite being in the top-performing sector, agriculture scored only 0.51 in 2018, and the electricity sector obtained 0.412. Results also show that even the best-performing sector operates below the efficiency level. The mining and quarrying sector ranked second by obtaining 0.623 EPI and 0.035 SBEPI. Results also show that much of the energy supply of Pakistan (60.17%) is focused on fossil fuels, supplemented by hydropower (33%), while nuclear, wind, biogas, and solar power account for 5.15%, 0.47%, 0.32%, and 0.03%, respectively. Nonetheless, the overall results for both measures remained reasonably consistent. According to the literature and the energy crisis and climate instability dilemma, the authors conclude that changes to a diverse green power network are a possibility and an imminent need. Similarly, the government should penalize companies with poor performance. Furthermore, to ensure the capacity development and stability of environmental management and associated actions in the country, providing access to knowledge and training to groom human resources and achieve the highest performance is crucial.

A Wirelessly Controlled Shape-Memory Alloy-Based Bistable Metal Swimming Device

Shape memory Nitinol has long been used for actuation. However, utilizing Nitinol to fabricate novel devices for various applications is a challenge, but has shown incredible promise and impacts. Bistable metal strips are widely adopted for shape morphing purposes (primarily in kid’s toys, e.g., snap bracelets) due to their easy and robust transformation between two states. In this paper, we combine Nitinol shape memory alloy and bistable metal strip to fabricate a swimming actuator with both slow moving and fast snapping capability, akin to an octopus swimming slowly in water, but quickly moving upon encountering a threat. The actuator developed here can also swim in multiple directions, all controlled by a wireless module. Furthermore, we demonstrate that an on-board sensor can be incorporated for potential environmental monitoring applications. Taken together, along with the fact that the device developed here has no mechanical parts, makes this an interesting potential alternative to more expensive, and energy consuming boats.

Comparative Study of Metakaolin and Zeolite Tuff Influence on Properties of High-Strength Concrete

Composite admixtures which include active pozzolanic components and high-range water reducers, allows to obtain high-strength, particularly dense and durable concrete to achieve a reduction in resources and energy consumption of manufacturing.Zeolite, containing a significant amount of active silica, can serve as one of the alternative substances to resources and energy consuming mineral admixtures like metakaolin and silica fume. The deposits of zeolites are developed in Transcarpathia (Ukraine), USA, Japan, New Zealand, Iceland and other countries. It is known that zeolite tuffs exhibit pozzolanic properties and are capable to substitution reactions with calcium hydroxide.However, the high dispersion of zeolite rocks leads to a significant increase in the water consumption of concrete. Simultaneous introduction of zeolite tuffs with superplasticizers, which significantly reduce the water content, creates the preconditions for their effective use in high-strength concrete.Along with dehydrated (calcined) zeolite, natural (non-calcined) zeolite expresses itself as an effective mineral admixture of concrete. When using non-calcined zeolite, the effect of increasing in compressive strength at the age of 3 and 7 days is close to the effect obtained when using dehydrated zeolite: 8-10% and 10- 12%, respectively, and 28 days the strength growth is 13-22%. The use of non-calcined zeolite has a significant economic feasibility, so it certainly deserves attention. There were compared the effect of zeolite to metakaolinThe results of the research indicate that the use of composite admixtures, consisted of calcined (non-calcined) zeolite tuff of high dispersity and superplasticizer of naphthalene formaldehyde type, allows to obtain concretes classes C50…C65.

Deep Learning for Photonic Design and Analysis: Principles and Applications

Innovative techniques play important roles in photonic structure design and complex optical data analysis. As a branch of machine learning, deep learning can automatically reveal the inherent connections behind the data by using hierarchically structured layers, which has found broad applications in photonics. In this paper, we review the recent advances of deep learning for the photonic structure design and optical data analysis, which is based on the two major learning paradigms of supervised learning and unsupervised learning. In addition, the optical neural networks with high parallelism and low energy consuming are also highlighted as novel computing architectures. The challenges and perspectives of this flourishing research field are discussed.

The CovRS Environmental Sensor Directly Controls the ComRS Signaling System To Orchestrate Competence Bimodality in Salivarius Streptococci

Combining production of antibacterial compounds and uptake of DNA material released by dead cells, competence is one of the most efficient survival strategies in streptococci. Yet, this powerful tactic is energy consuming and reprograms the metabolism to such an extent that cell proliferation is transiently impaired.

FCERP: A Novel WSNs Fuzzy Clustering and Energy Efficient Routing Protocol

Wireless sensor networks (WSNs) are set of sensor nodes to monitor and detect transmitted data to the sink. WSNs face significant challenges in terms of node energy availability, which may impact network sustainability. As a result, developing protocols and algorithms that make the best use of limited resources, particularly energy resources, is critical issues for designing WSNs. Routing algorithms, for example, are unique algorithms as they have a direct and effective relationship with lifetime of network and energy. The available routing protocols employ single-hop data transmission to the sink and clustering per round. In this paper, a Fuzzy Clustering and Energy Efficient Routing Protocol (FCERP) that lower the WSNs energy consuming and increase the lifetime of network is proposed. FCERP introduces a new cluster-based fuzzy routing protocol capable of utilizing clustering and multiple hop routing features concurrently using a threshold limit. A novel aspect of this research is that it avoids clustering per round while considering using fixed threshold and adapts multi-hop routing by predicting the best intermediary node for clustering and the sink. Some Fuzzy factors such as residual energy, neighbors amount, and distance to sink considered when deciding which intermediary node to use.

FABRICATION CYCLES COMPARISON OF ASSEMBLIES AND MONOLITHIC PARTS MADE BY 3D PRINTING METHOD

The article presents an analysis of the time-consuming, energy-consuming, and cost-consuming nature of 3D printing a three-dimensional polymer components made in two separate approaches: assembly and monolith structure of various materials (automatic filament change required). The introduction includes the definition of 3D printing, its advantages and examples of practical applications, as well as the reason for undertaking the researches described in the article. The justification of the form of 3D sample models was discussed in detail, as well as the methodology adopted by the authors for comparing the print characteristics and the steps of the printing cycles (print preparation, the course of the printing process and post-processing). A comparison of the materials consumption in the phasess of manual and automatic filament change in the mixer were also described. The test printout was made on the Prusa i3 MK3S printer for filament deposition (FDM or FFF methods). For automatic filament mixing, the Palette 2 Pro device was used. The conclusions also include guidelines for the design and production of models made in one continuous printing cycle (using automatic filament feeding devices). Monolithic elements are less accurate, while elements with replaceable filaments are cheaper, less energy-consuming and the material consumption is lower.

A Method for the Segregation of Emulsion Inner Phase Droplets Using Imbibition Process in Porous Material

The process of forming an emulsion is an energy-consuming process. The smaller the internal phase droplets we want to produce and the closer the droplets are in size to each other (monodisperse), the more energy we need to put into the system. Generating energy carries a high economic cost, as well as a high environmental footprint. Considering the fact that dispersive systems are widely used in various fields of life, it is necessary to search for other, less-energy-intensive methods that will allow the creation of dispersive systems with adequate performance and minimal energy input. Therefore, an alternative way to obtain emulsions characterized by small droplet sizes was proposed by using an imbibition process in porous materials. By applying this technique, it was possible to obtain average droplet sizes at least half the size of the base emulsion while reducing the polydispersity by about 40%. Oil-in-water emulsions in which vegetable oil or kerosene is the oily phase were tested. The studies were carried out at three different volume concentrations of the emulsions. Detailed analyses of diameter distributions and emulsion concentrations are presented. In addition, the advantages and limitations of the method are presented and the potential for its application is indicated.