Computational Power Research Articles

Multi-Agent Collaborative Bayesian Optimization via Constrained Gaussian Processes

The increase in the computational power of edge devices has opened a new paradigm for collaborative analytics whereby agents borrow strength from each other to improve their learning capabilities. This work focuses on collaborative Bayesian optimization (BO), in which agents work together to efficiently optimize black-box functions without the need for sensitive data exchange. Our idea revolves around introducing a class of constrained Gaussian process surrogates, enabling agents to borrow informative designs from high-performing collaborators to enhance and expedite their optimization process. Our approach presents the first general-purpose collaborative BO framework that is compatible with any Gaussian process kernel and most of the known acquisition functions. Despite the simplicity of our approach, we demonstrate that it offers elegant theoretical guarantees and significantly outperforms state-of-the-art methods, especially when agents have heterogeneous black-box functions. Through various simulations and a real-life experiment in additive manufacturing, we showcase the advantageous properties of our approach and the benefits derived from collaboration.

Sci-tech Finance and Technological Innovation of AI Computing Power Enterprises: Based on the Perspective of Financial Regulation

As the new round of technological revolution continues to evolve, computing power is progressively emerging as the "new qualitative productive force" in the era of digital economy; AI computing power, being the core component of computing power, has consequently garnered extensive attention. This study examines the impact of sci-tech finance on the technological innovation of AI computing power enterprises, the mechanism of action, and the role of financial regulation, using data from Chinese A-share listed AI computing power enterprises from 2014 to 2022. The findings are as follows: Firstly, the development of sci-tech finance significantly enhances the technological innovation capacity of AI computing power enterprises. Secondly, the alleviation of financing constraints partially mediates this effect. Thirdly, overall financial regulation exerts a negative regulatory effect on the relationship between sci-tech finance and technological innovation; specifically, financial regulation in the eastern region has a negative regulatory role, whereas in the non-eastern regions, it exhibits a positive regulatory role. These research conclusions hold certain implications and reference value for enhancing the sci-tech finance system, fostering the development of the computing power industry, and reforming the regulatory system.

Robot Motion Planning Based on an Adaptive Slime Mold Algorithm and Motion Constraints

The rapid advancement of artificial intelligence technology has significantly enhanced the intelligence of mobile robots, facilitating their widespread utilization in unmanned driving, smart home systems, and various other domains. As the scope, scale, and complexity of robot deployment continue to expand, there arises a heightened demand for enhanced computational power and real-time performance, with path planning emerging as a prominent research focus. In this study, we present an adaptive Lévy flight–rotation slime mold algorithm (LRSMA) for global robot motion planning, which incorporates LRSMA with the cubic Hermite interpolation. Unlike traditional methods, the algorithm eliminates the need for a priori knowledge of appropriate interpolation points. Instead, it autonomously detects the convergence status of LRSMA, dynamically increasing interpolation points to enhance the curvature of the motion curve when it surpasses the predefined threshold. Subsequently, it compares path lengths resulting from two different objective functions to determine the optimal number of interpolation points and the best path. Compared to LRSMA, this algorithm reduced the minimum path length and average processing time by (2.52%, 3.56%) and (38.89%, 62.46%), respectively, along with minimum processing times. Our findings demonstrate that this method effectively generates collision-free, smooth, and curvature-constrained motion curves with the least processing time.

FastFaceCLIP: A lightweight text‐driven high‐quality face image manipulation

AbstractAlthough many new methods have emerged in text‐driven images, the large computational power required for model training causes these methods to have a slow training process. Additionally, these methods consume a considerable amount of video random access memory (VRAM) resources during training. When generating high‐resolution images, the VRAM resources are often insufficient, which results in the inability to generate high‐resolution images. Nevertheless, recent Vision Transformers (ViTs) advancements have demonstrated their image classification and recognition capabilities. Unlike the traditional Convolutional Neural Networks based methods, ViTs have a Transformer‐based architecture, leverage attention mechanisms to capture comprehensive global information, moreover enabling enhanced global understanding of images through inherent long‐range dependencies, thus extracting more robust features and achieving comparable results with reduced computational load. The adaptability of ViTs to text‐driven image manipulation was investigated. Specifically, existing image generation methods were refined and the FastFaceCLIP method was proposed by combining the image‐text semantic alignment function of the pre‐trained CLIP model with the high‐resolution image generation function of the proposed FastFace. Additionally, the Multi‐Axis Nested Transformer module was incorporated for advanced feature extraction from the latent space, generating higher‐resolution images that are further enhanced using the Real‐ESRGAN algorithm. Eventually, extensive face manipulation‐related tests on the CelebA‐HQ dataset challenge the proposed method and other related schemes, demonstrating that FastFaceCLIP effectively generates semantically accurate, visually realistic, and clear images using fewer parameters and less time.

Analysis of the Use of Artificial Intelligence in Software-Defined Intelligent Networks: A Survey

The distributed structure of traditional networks often fails to promptly and accurately provide the computational power required for artificial intelligence (AI), hindering its practical application and implementation. Consequently, this research aims to analyze the use of AI in software-defined networks (SDNs). To achieve this goal, a systematic literature review (SLR) is conducted based on the PRISMA 2020 statement. Through this review, it is found that, bottom-up, from the perspective of the data plane, control plane, and application plane of SDNs, the integration of various network planes with AI is feasible, giving rise to Intelligent Software Defined Networking (ISDN). As a primary conclusion, it was found that the application of AI-related algorithms in SDNs is extensive and faces numerous challenges. Nonetheless, these challenges are propelling the development of SDNs in a more promising direction through the adoption of novel methods and tools such as route optimization, software-defined routing, intelligent methods for network security, and AI-based traffic engineering, among others.

An exhaustive review of the stream ciphers and their performance analysis

The number of internet of things (IoT) applications has increased, which has increased the demand for low-resource gadgets. The data produced by these devices must be protected to guarantee security. The devices operate in conditions with limited space, computational power, memory, and energy. High-security standards are difficult to achieve with limited resources. The detailed analysis of various stream ciphers and their performance metrics is reviewed in this manuscript. The functionality of the stream ciphers is categorized and thoroughly discussed based on both the hardware and software viewpoints. The security attacks and their countermeasure methods using stream ciphers are discussed. The performance metrics of most hardware-based stream ciphers, including the ECRYPT stream cipher project (eSTREAM) ciphers, are discussed. Each hardware stream cipher design highlights the hardware constraints such as chip area, frequency, throughput, and hardware efficiency. The work also highlights the various applications using these stream ciphers. The current trends using these stream ciphers are discussed with futuristic goals.

Artificial intelligence-powered intelligent reflecting surface systems countering adversarial attacks in machine learning

With the increase in the computation power of devices wireless communication has started adopting machine learning (ML) techniques. Intelligent reflecting surface (IRS) is a programmable device that can be used to control electromagnetic wave propagation by changing the electric and magnetic values of its surface. State-of-the-art ML especially on deep learning (DL)-based IRS-enhanced communication is an emerging topic. Yet while integrating IRS with other emerging technologies possibilities of adversarial data creaping is high. Threats to security, their mitigation, and complexes for AI-powered applications in next generation networks are continuously emerging. In this work the ability of an IRS enhanced wireless network in future-generation networks to prevent adversarial machinelearning attacks is studied. The artificial intelligence (AI) model is used to minimize the susceptibility of attacks using defense distillation mitigation technique. The outcome shows that the defensive distillation technique (DDT) increases the strength and performance by around 22% of the AI method under an adversarial attack.

Deep learning for economic transformation: a parametric review

Deep learning (DL) is increasingly recognized for its effectiveness in analyzing and forecasting complex economic systems, particularly in the context of Pakistan's evolving economy. This paper investigates DL's transformative role in managing and interpreting increasing volumes of intricate economic data, leading to more nuanced insights. DL models show a marked improvement in predictive accuracy and depth over traditional methods across various economic domains and policymaking scenarios. Applications include demand forecasting, risk evaluation, market trend analysis, and resource allocation optimization. These processes utilize extensive datasets and advanced algorithms to identify patterns that traditional methods cannot detect. Nonetheless, DL's broader application in economic research faces challenges like limited data availability, complexity of economic interactions, interpretability of model outputs, and significant computational power requirements. The paper outlines strategies to overcome these barriers, such as enhancing model interpretability, employing federated learning for better data privacy, and integrating behavioral and social economic theories. It concludes by stressing the importance of targeted research and ethical considerations in maximizing DL's impact on economic insights and innovation, particularly in Pakistan and globally.

MDDBranchNet: A Deep Learning Model for Detecting Major Depressive Disorder Using ECG Signal.

Major depressive disorder (MDD) is a chronic mental illness which affects people's well-being and is often detected at a later stage of depression with a likelihood of suicidal ideation. Early detection of MDD is thus necessary to reduce the impact, however, it requires monitoring vitals in daily living conditions. EEG is generally multi-channel and due to difficulty in signal acquisition, it is unsuitable for home-based monitoring, whereas, wearable sensors can collect single-channel ECG. Classical machine-learning based MDD detection studies commonly use various heart rate variability features. Feature generation, which requires domain knowledge, is often challenging, and requires computation power, often unsuitable for real time processing, MDDBranchNet is a proposed parallel-branch deep learning model for MDD binary classification from a single channel ECG which uses additional ECG-derived signals such as R-R signal and degree distribution time series of horizontal visibility graph. The use of derived branches was able to increase the model's accuracy by around 7%. An optimal 20-second overlapped segmentation of ECG recording was found to be beneficial with a 70% prediction threshold for maximum MDD detection with a minimum false positive rate. The proposed model evaluated MDD prediction from signal excerpts, irrespective of location (first, middle or last one-third of the recording), instead of considering the entire ECG signal with minimal performance variation stressing the idea that MDD phenomena are likely to manifest uniformly throughout the recording.

MCU-based Safer Coagulation Mode by Nonfixed Duty Cycle for an Electrosurgery Inverter.

The arcing-involved pulsating coagulation mode with both active and blank periods is essential for modern electrosurgery. This paper begins with a comprehensive introduction to such a pulsating mode, followed by its implementation challenges. Then, an industrial-scale low-speed microcontroller unit (MCU), TMS320F28379D, is utilized to exemplify the proposed output sampling and data-transferring strategy on a gallium nitride (GaN)-based high-frequency inverter that enables coagulation mode with interweaved active periods and blank periods. The inverter prototype fills the active period with 390 kHz sinusoids of amplitude ranging from hundreds to thousands of Volts, while maintaining null outputs during blank periods. The strategy of sampling the above-mentioned sinusoidal outputs, coupled with their data transfer facilitated by direct memory access (DMA), is also articulated for subsequential power computation. Besides that, a novel nonfixed duty cycle approach, featuring an alterable number of sinusoids as the active period, is proposed and integrated into the GaN-based inverter to enhance mode safety. Finally, the power tracking performance of the mode is evaluated initially on resistive load, secondarily on resistive plus capacitive load (R-C), and thirdly on fresh biotissue with the appearance of electrical arcing. The existing necessity of the null blank periods is examined at the end of the paper.

개념 디자인에서의 AI의 활용에 관한 연구

Machine learning(ml) was already developed in the 1940s, but its application value was low due to limited computational power and lack of data. Since then, based on advancements in computer technology and data accumulation, artificial intelligence, particularly machine learning, has gained interdisciplinary attention as a tool for providing insights in the field of design. Recently, numerous prior studies on image generation using artificial intelligence have emerged, reflecting this growing interest. However, from a practical standpoint, there are several criticisms: the images generated by machine learning often lack the resolution required by designers, and they fall short in providing original images. It is worth examining whether the perspectives of these prior studies are becoming teleologically biased. Therefore, this study aims to explore the gap between interdisciplinary discussions and practical applications, seeking to bridge this gap and integrate AI-generated image creation into practical use. This research is based on a literature review of Generative Adversarial Networks(GANs) in machine learning(ml), examining whether the images generated are acceptable from the perspective of product designers. Based on this, we will use text mining methods to perform clustering, information extraction, and sentiment analysis to understand consumer attitudes. Through this approach, we aim to explore the potential use of AI in product design from the standpoint of consumers and practitioners rather than an interdisciplinary perspective.

Molecular Dynamics Simulations Help Determine the Molecular Mechanisms of Lasioglossin-III and Its Variant Peptides' Membrane Interfacial Interactions.

Lasioglossin-III (LL-III) is a potent broad-spectrum antimicrobial peptide used in diverse antimicrobial applications. In this work, coarse-grained and all-atom molecular dynamics simulation strategies were used in tandem to interpret the molecular mechanisms involved in the interfacial dynamics of LL-III and its recombinant variants during interactions with diverse cell membrane systems. Our results indicate that the membrane charges act as the driving force for initiating the membrane-peptide interactions, while the hydrophobic or van der Waals forces help to reinforce the membrane-peptide bindings. The optimized charge-hydrophobicity ratio of the LL-III peptides helps ensure their high specificity toward bacterial membranes compared to mammalian membrane systems, which also helps explain our experimental observations. Overall, we hope that our work gives new insight into the antimicrobial action of LL-III peptides and that the adopted simulation strategy will help other scientists and engineers extract maximal information from complex molecular simulations using minimal computational power.

Data and Energy Impacts of Intelligent Transportation—A Review

The deployment of intelligent transportation is still in its early stages and there are many challenges that need to be addressed before it can be widely adopted. Autonomous vehicles are a class of intelligent transportation that is rapidly developing, and they are being deployed in selected cities. A combination of advanced sensors, machine learning algorithms, and artificial intelligence are being used in these vehicles to perceive their environment, navigate, and make the right decisions. These vehicles leverage extensive data sourced from various sensors and computers integrated into the vehicle. Hence, massive computational power is required to process the information from various built-in sensors in milliseconds to make the right decision. The power required by the sensors and the use of additional computational power increases the energy consumption, and, hence, could reduce the range of the autonomous electric vehicle relative to a standard electric car and lead to additional emissions. A number of review papers have highlighted the environmental benefits of autonomous vehicles, focusing on aspects like optimized driving, improved route selection, fewer stops, and platooning. However, these reviews often overlook the significant energy demands of the hardware systems—such as sensors, computers, and cameras—necessary for full autonomy, which can decrease the driving range of electric autonomous vehicles. Additionally, previous studies have not thoroughly examined the data processing requirements in these vehicles. This paper provides a more detailed review of the volume of data and energy usage by various sensors and computers integral to autonomous features in electric vehicles. It also discusses the effects of these factors on vehicle range and emissions. Furthermore, the paper explores advanced technologies currently being developed by various industries to enhance processing speeds and reduce energy consumption in autonomous vehicles.

Classification of white blood cells (leucocytes) from blood smear imagery using machine and deep learning models: A global scoping review.

Machine learning (ML) and deep learning (DL) models are being increasingly employed for medical imagery analyses, with both approaches used to enhance the accuracy of classification/prediction in the diagnoses of various cancers, tumors and bloodborne diseases. To date however, no review of these techniques and their application(s) within the domain of white blood cell (WBC) classification in blood smear images has been undertaken, representing a notable knowledge gap with respect to model selection and comparison. Accordingly, the current study sought to comprehensively identify, explore and contrast ML and DL methods for classifying WBCs. Following development and implementation of a formalized review protocol, a cohort of 136 primary studies published between January 2006 and May 2023 were identified from the global literature, with the most widely used techniques and best-performing WBC classification methods subsequently ascertained. Studies derived from 26 countries, with highest numbers from high-income countries including the United States (n = 32) and The Netherlands (n = 26). While WBC classification was originally rooted in conventional ML, there has been a notable shift toward the use of DL, and particularly convolutional neural networks (CNN), with 54.4% of identified studies (n = 74) including the use of CNNs, and particularly in concurrence with larger datasets and bespoke features e.g., parallel data pre-processing, feature selection, and extraction. While some conventional ML models achieved up to 99% accuracy, accuracy was shown to decrease in concurrence with decreasing dataset size. Deep learning models exhibited improved performance for more extensive datasets and exhibited higher levels of accuracy in concurrence with increasingly large datasets. Availability of appropriate datasets remains a primary challenge, potentially resolvable using data augmentation techniques. Moreover, medical training of computer science researchers is recommended to improve current understanding of leucocyte structure and subsequent selection of appropriate classification models. Likewise, it is critical that future health professionals be made aware of the power, efficacy, precision and applicability of computer science, soft computing and artificial intelligence contributions to medicine, and particularly in areas like medical imaging.

Fisher’s pioneering work on discriminant analysis and its impact on Artificial Intelligence

Sir Ronald Aylmer Fisher opened many new areas in Multivariate Analysis, and the one which we will consider is discriminant analysis. Several papers by Fisher and others followed from his seminal paper in 1936 where he coined the name discrimination function. Historically, his four papers on discriminant analysis during 1936–1940 connect to the contemporaneous pioneering work of Hotelling and Mahalanobis. We revisit the famous iris data which Fisher used in his 1936 paper and in particular, test the hypothesis of multivariate normality for the data which he assumed. Fisher constructed his genetic discriminant motivated by this application and we provide a deeper insight into this construction; however, this construction has not been well understood as far as we know. We also indicate how the subject has developed along with the computer revolution, noting newer methods to carry out discriminant analysis, such as kernel classifiers, classification trees, support vector machines, neural networks, and deep learning. Overall, with computational power, the whole subject of Multivariate Analysis has changed its emphasis but the impact of this Fisher’s pioneering work continues as an integral part of supervised learning in Artificial Intelligence (AI).

Tricking AI chips into simulating the human brain: A detailed performance analysis

In recent years, significant strides in Artificial Intelligence (AI) have led to various practical applications, primarily centered around training and deployment of deep neural networks (DNNs). These applications, however, require considerable computational resources, predominantly reliant on modern Graphics-Processing Units (GPUs). Yet, the quest for larger and faster DNNs has spurred the creation of specialized AI chips and efficient Machine-Learning (ML) software tools like TensorFlow and PyTorch have been developed for striking a balance between usability and performance. Simultaneously, the field of computational neuroscience shares a similar quest for increased computational power to simulate more extensive and detailed brain models, while also keeping usability high. Although GPUs have also entered this field, programming complexity remains high, resulting in cumbersome simulations. Inspired by AI progress, we introduce a workflow for easily accelerating brain simulations using TensorFlow and evaluate the performance of various, cutting-edge AI chips – including the Graphcore Intelligence-Processing Unit (IPU), GroqChip, Nvidia GPU with Tensor Cores, and Google Tensor-Processing Unit (TPU) – when simulating a biologically detailed as well as simpler brain models. Our model simulations explore the architectural tradeoffs of a modern-day CPU and these four AI platforms by varying computational density, memory requirements and floating-point numerical accuracy. Results show that the GroqChip achieves the best performance for small networks, yet is unable to simulate large-scale networks. At the scale of mammalian brains, the GPU, IPU and TPU achieve speedups ranging from 29x to 1,208x times over CPU runtimes. Remarkably, the TPU sets a new record for the largest, real-time simulation of the inferior-olivary nucleus in the brain. Reduced-accuracy floating-point implementations make some simulation results unreliable for brain research, notably for the GroqChip. Consequently, this work underscores the potential of ML libraries for accelerating brain simulations as well as the critical role of AI-chip numerical accuracy for biophysically realistic brain models.

Next-Generation Patient-Based Real-Time Quality Control Models.

Patient-based real-time QC (PBRTQC) uses patient-derived data to assess assay performance. PBRTQC algorithms have advanced in parallel with developments in computer science and the increased availability of more powerful computers. The uptake of Artificial Intelligence in PBRTQC has been rapid, with many stated advantages over conventional approaches. However, until this review, there has been no critical comparison of these. The PBRTQC algorithms based on moving averages, regression-adjusted real-time QC, neural networks and anomaly detection are described and contrasted. As Artificial Intelligence tools become more available to laboratories, user-friendly and computationally efficient, the major disadvantages, such as complexity and the need for high computing resources, are reduced and become attractive to implement in PBRTQC applications.

Does a higher hashrate strengthen Bitcoin network security?

In the blockchain world, proof-of-work is the dominant protocol mechanism that determines the consensus of the ledger. The hashrate, a measure of the computational power directed toward securing a blockchain through proof-of-work consensus, is a fundamental measure of preventing various attacks. This study tests the causal relationship between the hashrate and the security outcome of the Bitcoin blockchain. We use vector error correction modeling to analyze the endogenous relationships between the hashrate, Bitcoin price, and transaction fee, revealing the need for an additional variable to achieve our aim. Employing a measure summarizing the growth of demand factors in the Bitcoin ecosystem indicates that hashrate fluctuations significantly influence security level changes. This result underscores the importance of the hashrate in ensuring the security of the Bitcoin blockchain.

IMAGE CLASSIFIER FOR FAST SEARCH IN LARGE DATABASES

Relevance. The avalanche-like growth in the amount of information on the Internet necessitates the development of effective methods for quickly processing such information in information systems. Clustering of news information is carried out by taking into account both the morphological analysis of texts and graphic content. Thus, an urgent task is the clustering of images accompanying textual information on various web resources, including news portals. The subject of study is an image classifier that exhibits low sensitivity to increased information in databases. The purpose of the article is to enhance the efficiency of searching for identical images in databases experiencing a daily influx of 10-12 thousand images, by developing an image classifier. Methods used: mathematical modeling, content-based image retrieval, two-dimensional discrete cosine transform, image processing methods, decision-making methods. The following results were obtained. An image classifier has been developed with low sensitivity to increased database information. The properties of the developed classifier have been analyzed. The experiments demonstrated that clustering information based on images using the developed classifier proved to be sufficiently fast and cost-effective in terms of information volumes and computational power requirements.

Estimation of bandwidth, voltage swing and DC coefficients for VLSI interconnects: current mode technique

ABSTRACT Necessity of bandwidth requires the effective use of it for high speed data transmission and faster read/write operation in memory cells. In this paper the transfer function expressions for RC interconnect line is presented for resistive load. The line delay, bandwidth and step response for current and voltage mode signalling is formulated. The current mode and voltage mode signalling provides the 3-dB bandwidth of 6.8 GHz and 3.1 GHz, respectively. An interconnect line modelled as either RLC or RC line depending upon input signal rise time, damping, and range of interconnect length with significant inductance effect is investigated for global interconnect lines. Low swing for column bit line in memory architectures is very important and estimated in terms of dc coefficients. For RC global interconnect of 10 mm, a swing of 0.382 V is obtained for current mode signalling, when a power supply of 1.8 V is applied to the network. DC coefficient and steady state value has been obtained which is significant for power computation on column bit line. It is also analytically validated by SPICE and MATLAB simulations.