Parallel Techniques Research Articles

Comparative evaluation and patient satisfaction with an electrical impedance-based device versus digital radiography in the estimation of remaining dentin thickness in carious posterior permanent teeth: (Diagnostic accuracy study)

BackgroundAccurate assessment of remaining dentin thickness (RDT) is paramount for restorative decisions and treatment planning of vital teeth to avoid any pulpal injury. This diagnostic accuracy study compared the validity and patient satisfaction of an electrical impedance based device Prepometer™ (Hager & Werken, Duisburg, Germany) versus intraoral digital radiography for the estimation of remaining dentin thickness in carious posterior permanent teeth.MethodsSeventy patients aged 12–25 years with carious occlusal or proximal permanent vital posterior teeth were recruited. Tooth preparation was performed to receive an adhesive restoration. Pre- and post-excavation RDT were measured radiographically by two calibrated raters using the paralleling periapical technique. Prepometer™ measurements were performed by the operator. Patients rated their satisfaction level with each tool on a 4-point Likert scale and 100 mm visual analog scale (VAS). Inter and intragroup comparisons were analyzed using signed rank test, while agreement between devices and observations was tested using weight kappa (WK) coefficient.Resultsthe intergroup comparisons showed that, before and after excavation, there was a significant difference between measurements made by both techniques (p < 0.001). After excavation, there was a weak agreement between measurements (WK = 0.2, p < 0.001), whereas before excavation, the agreement was not statistically significant (p = 0.407). Patients were significantly more satisfied with Prepometer™ based on scales and VAS (p < 0.001).ConclusionPrepometer™ could be a viable clinical tool for determining RDT with high patient satisfaction, while radiographs tended to overestimate RDT in relation to the Prepometer™.

Optimizing DNN training with pipeline model parallelism for enhanced performance in embedded systems

Deep Neural Networks (DNNs) have gained widespread popularity in different domain applications due to their dominant performance. Despite the prevalence of massively parallel multi-core processor architectures, adopting large DNN models in embedded systems remains challenging, as most embedded applications are designed with single-core processors in mind. This limits DNN adoption in embedded systems due to inefficient leveraging of model parallelization and workload partitioning. Prior solutions attempt to address these challenges using data and model parallelism. However, they lack in finding optimal DNN model partitions and distributing them efficiently to achieve improved performance.This paper proposes a DNN model parallelism framework to accelerate model training by finding the optimal number of model partitions and resource provisions. The proposed framework combines data and model parallelism techniques to optimize the parallel processing of DNNs for embedded applications. In addition, it implements the pipeline execution of the partitioned models and integrates a task controller to manage the computing resources. The experimental results for image object detection demonstrate the applicability of our proposed framework in estimating the latest execution time and reducing overall model training time by almost 44.87% compared to the baseline AlexNet convolutional neural network (CNN) model.

Graphics processing unit (GPU)-enhanced nonhydrostatic model with grid nesting for global tsunami propagation and coastal inundation

Nonhydrostatic models have proven their superiority in describing tsunami propagation over trans-oceanic distances and nearshore transformation because of their good dispersion and nonlinearity properties. The novel one-layer nonhydrostatic formulations proposed by Wang et al. [Phys. Fluids 35, 076610 (2023)] have been rederived in the spherical coordinate system incorporating Coriolis effects to enable the application of basin-wide tsunami propagation. The model was implemented using the fractional step method, where the hydrostatic step was solved by a Godunov-type finite-volume scheme, and the nonhydrostatic step was obtained with the finite-difference method. Additionally, a two-way grid-nesting scheme was employed to adapt the topographic features for efficient computation of tsunami propagation in deep ocean and coastal inundation. Furthermore, graphics processing unit (GPU)-parallelism technique was incorporated to further optimize the model performance. An idealized benchmark test as well as three experiments of regular and irregular waves, solitary, and N-waves transformations have been simulated to demonstrate the superior performance of the current GPU-accelerated grid-nesting nonhydrostatic model. Finally, the model has been applied to reproduce the 1964 Prince William Sound Tsunami, its propagation across the North Pacific and induced inundation in the Seaside.

Comparative examination of trabecular bone structure in healthy and diabetic patients with periodontitis using fractal analysis.

Fractal analysis is a numerical method that indicates the structural patterns and complexity of the trabecular bone on radiographs. The aim of this cross-sectional study was to evaluate the trabecular bone structure in systemically healthy patients and diabetes mellitus patients with periodontitis using fractal analysis. The study included 125 mandibular first molars of nonsmoker patients. The subjects were divided into five subgroups: diabetes mellitus patients with mild-moderate periodontitis, diabetes mellitus patients with advanced periodontitis, systemically healthy individuals with mild-moderate periodontitis, systemically healthy individuals with advanced periodontitis, and systemically healthy individuals with gingivitis (control group). Clinical periodontal parameters (pocket depth, bleeding on probing, clinical attachment loss, and bone loss) were recorded. Two specific sites located in the mesial-distal regions (n = 250) of the mandibular first molars were identified using periapical radiographs captured with a parallel technique. Fractal analysis values were calculated using the box-counting method. One-way analysis of variance (ANOVA), and Pearson correlation analysis were used for statistical evaluation. The highest fractal analysis values were observed in systemically healthy with gingivitis patients (mesial fractal analysis: 1.86 ± 0.01; distal fractal analysis: 1.85 ± 0.01). Patients with periodontitis (mesial fractal analysis: 1.78 ± 0.02; distal fractal analysis: 1.79 ± 0.01) exhibited lower fractal analysis values compared to the control group. There was no significant difference in mesial and distal fractal analysis values between all periodontitis groups. No correlation was found between age, sex, clinical attachment loss, bone loss, and fractal analysis (P > .05). Although fractal analysis values were lower in the periodontitis groups compared to the control group, fractal analysis did not demonstrate any periodontitis-associated changes of bone trabeculation in diabetes at any stage of periodontitis. Furthermore, there was no significant association between fractal analysis values and age, sex, clinical attachment, and bone loss.

Accessible luminal interface of bovine rectal organoids generated from cryopreserved biopsy tissues.

Developing precise species-specific in vitro models that closely resemble in vivo intestinal tissues is essential for advancing our understanding of gastrointestinal physiology and associated diseases. This is especially crucial in examining host-pathogen interactions, particularly in bovines, a known reservoir for microbes and pathogens posing substantial public health threats. This research investigated the viability of producing bovine rectal organoids from cryopreserved tissues. We compared two cryopreservation methods with a traditional technique using fresh tissues, evaluating their effectiveness through growth rates, long-term viability, and comprehensive structural, cellular, and genetic analyses. These assessments utilized phase-contrast imaging, immunofluorescence imaging, and RT-qPCR assays. Additionally, the study developed a sophisticated method for forming a functional epithelial barrier from organoid-derived bovine rectal monolayers, incorporating a wide range of epithelial cells. This methodology employed transepithelial electrical resistance (TEER), parallel artificial membrane permeability assay (Papp), confocal microscopy, and advanced imaging techniques like scanning electron microscopy (SEM) and transmission electron microscopy (TEM). Our findings decisively show that bovine rectal organoids can be effectively generated from cryopreserved biopsy tissues. Moreover, we formulated a robust and optimized protocol for creating functional rectal monolayers from these organoids. This significant progress is particularly relevant given the susceptibility of the bovine rectum to various enteric pathogens of public health concern, marking a vital step forward in veterinary and biomedical research. The creation of accurate species specific in vitro models that faithfully mimic in vivo intestinal tissues is critical for enhancing our understanding of gut physiology and related pathologies. This is particularly relevant in studying the interactions between hosts and microbes or pathogens with significant public health risks where bovine can be the major reservoir.

Mathematical modeling and problem solving: from fundamentals to applications

The rapidly advancing fields of machine learning and mathematical modeling, greatly enhanced by the recent growth in artificial intelligence, are the focus of this special issue. This issue compiles extensively revised and improved versions of the top papers from the workshop on Mathematical Modeling and Problem Solving at PDPTA'23, the 29th International Conference on Parallel and Distributed Processing Techniques and Applications. Covering fundamental research in matrix operations and heuristic searches to real-world applications in computer vision and drug discovery, the issue underscores the crucial role of supercomputing and parallel and distributed computing infrastructure in research. Featuring nine key studies, this issue pushes forward computational technologies in mathematical modeling, refines techniques for analyzing images and time-series data, and introduces new methods in pharmaceutical and materials science, making significant contributions to these areas.

A new thread-level speculative automatic parallelization model and library based on duplicate code execution

Loop-efficient automatic parallelization has become increasingly relevant due to the growing number of cores in current processors and the programming effort needed to parallelize codes in these systems efficiently. However, automatic tools fail to extract all the available parallelism in irregular loops with indirections, race conditions or potential data dependency violations, among many other possible causes. One of the successful ways to automatically parallelize these loops is the use of speculative parallelization techniques. This paper presents a new model and the corresponding C++ library that supports the speculative automatic parallelization of loops in shared memory systems, seeking competitive performance and scalability while keeping user effort to a minimum. The primary speculative strategy consists of redundantly executing chunks of loop iterations in a duplicate fashion. Namely, each chunk is executed speculatively in parallel to obtain results as soon as possible and sequentially in a different thread to validate the speculative results. The implementation uses C++11 threads and it makes intensive use of templates and advanced multithreading techniques. An evaluation based on various benchmarks confirms that our proposal provides a competitive level of performance and scalability.

IoT-enabled Biometric Security: Enhancing Smart Car Safety with Depth-based Head Pose Estimation

Advanced Driver Assistance Systems (ADAS) are experiencing higher levels of automation, facilitated by the synergy among various sensors integrated within vehicles, thereby forming an Internet of Things (IoT) framework. Among these sensors, cameras have emerged as valuable tools for detecting driver fatigue and distraction. This study introduces HYDE-F, a Head Pose Estimation (HPE) system exclusively utilizing depth cameras. HYDE-F adeptly identifies critical driver head poses associated with risky conditions, thus enhancing the safety of IoT-enabled ADAS. The core of HYDE-F’s innovation lies in its dual-process approach: it employs a fractal encoding technique and keypoint intensity analysis in parallel. These two processes are then fused using an optimization algorithm, enabling HYDE-F to blend the strengths of both methods for enhanced accuracy. Evaluations conducted on a specialized driving dataset, Pandora, demonstrate HYDE-F’s competitive performance compared to existing methods, surpassing current techniques in terms of average Mean Absolute Error (MAE) by nearly 1 ∘ . Moreover, case studies highlight the successful integration of HYDE-F with vehicle sensors. Additionally, HYDE-F exhibits robust generalization capabilities, as evidenced by experiments conducted on standard laboratory-based HPE datasets, i.e., Biwi and ICT-3DHP databases, achieving an average MAE of 4.9 ∘ and 5 ∘ , respectively.

Implementation of FIR Filters through Inner product Units and Parallel Accumulations

Finite Impulse Response (FIR) filters are pivotal in digital signal processing, finding applications in diverse fields like audio processing, telecommunications, and biomedical signal analysis. This work presents an enhanced implementation methodology for FIR filters utilizing inner product computation and parallel accumulations. In the existing, FIR filters are typically implemented using convolution techniques, basic adders, and multipliers, which involve sequential processing and intensive computational resources. This method often leads to latency issues and limits real-time applications. Moreover, traditional implementations suffer from inefficiencies in utilizing hardware resources optimally, leading to suboptimal performance. The proposed methodology overcomes these limitations by leveraging inner product computations and parallel accumulation techniques. By exploiting inherent parallelism in the filtering process, the proposed method significantly reduces latency and enhances throughput.

Development of Distributed Search Algorithm for Improving Efficiency in Web Search Engines

This research article presents the development of a distributed search algorithm aimed at enhancing the efficiency of web search engines. Through a qualitative methodology involving literature review and library research, the study seeks to address the growing demand for improved search engine performance in handling large-scale data while maintaining fast response times. The literature review explores existing search algorithms and their limitations in processing vast amounts of information distributed across various servers. By analyzing prior research, the article identifies the need for a distributed approach to optimize search engine functionality and reduce latency. The developed distributed search algorithm employs parallel processing techniques to distribute search queries across multiple nodes, thereby improving overall system efficiency and response time. By leveraging distributed computing resources, the algorithm enhances the scalability and reliability of web search engines, enabling them to handle increased user traffic and data volume. The findings underscore the effectiveness of the proposed algorithm in significantly reducing search latency and improving search engine performance, as evidenced by experimental results. Furthermore, the article discusses potential applications of the distributed search algorithm in various domains, including e-commerce, information retrieval, and data analytics.

Parallel Fractional Stochastic Gradient Descent With Adaptive Learning for Recommender Systems

The structural change toward the digital transformation of online sales elevates the importance of parallel processing techniques in recommender systems, particularly in the pandemic and post-pandemic era. Matrix factorization (MF) is a popular and scalable approach in collaborative filtering (CF) to predict user preferences in recommender systems. Researchers apply Stochastic Gradient Descent (SGD) as one of the most famous optimization techniques for MF. Paralleling SGD methods help address Big Data challenges due to the wide range of products and the sparsity in user ratings. However, these methods&#x2019; convergence rate and accuracy are affected by the dependency between the user and item latent factors, specifically in large-scale problems. Besides, the performance is sensitive to the applied learning rates. This paper proposes a new parallel method to remove dependencies and boost speed-up by using fractional calculus to improve accuracy and convergence rate. We also apply adaptive learning rates to enhance the performance of our proposed method. The proposed method is based on Compute Unified Device Architecture (CUDA) platform. We evaluate the performance of our proposed method using real-world data and compare the results with the close baselines. The results show that our method can obtain high accuracy and convergence rate in addition to high parallelism.

Modified Broadband Ruthroff-Type Transmission Line Transformer Balun for Isolation-Enhanced Passive Mixer Design.

Generalized broadband operation facilitates multifunction or multiband highly integrated applications, such as modern transceiver systems, where ultra-wideband bidirectional passive mixers are favored to avoid a complex up/down-conversion scheme. In this paper, a modified Ruthroff-type transmission line transformer (TLT) balun is presented to enhance the isolation of the mixer from the local oscillator (LO) to the radio frequency (RF). Compared to the conventional methods, the proposed Ruthroff-type architecture adopts a combination of shunt capacitors and parallel coupled lines to improve the return loss at the LO port, thus effectively avoiding the area consumption for the diode-to-balun impedance transformation while simultaneously providing a suitable point for IF extraction. In addition, a parallel compensation technique consisting of an inductor and resistor is applied to the RF balun to significantly improve the amplitude/phase balance performance over a wide bandwidth. Benefiting from the aforementioned operations, an isolation-enhanced 8-30 GHz passive double-balanced mixer is designed as a proof-of-principle demonstration via 0.15-micrometer GaAs p-HEMT technology. It exhibits ultra-broadband performance with 7 dB average conversion loss and 50 dB LO-to-RF isolation under 15 dBm LO power. The monolithic microwave integrated circuit area is 0.96 × 1.68 mm2 including all pads.

KMS Infrastructure for KM Practice in Two Mobile Telecommunication Companies in Namibia

Knowledge management system infrastructure is at the forefront of knowledge management practice. Knowledge must be identified, captured, and shared to improve KM practices; however, little is known about the successful implementation of KMS infrastructure by Namibian mobile telecommunications companies. This mixed-methods research study employed a convergent parallel design and parallel sampling techniques. Three hundred and nine online questionnaires were distributed to a representative sample, with a response rate of 57%. Computer software packages were used to analyse quantitative data quantitatively, and qualitative data from interviews with 11 participants and analysis of documents were analysed thematically. Findings showed that infrastructure, management support, and employee participation were necessary for a successful KMS infrastructure implementation for effective and efficient KM practices. The study provides MT companies with a foundational understanding of how the infrastructure for KMS is essential for effective and efficient KM practice.

Investigating the applications and analysis of physics engine technologies

This research project delves into the performance impact of implementing parallel programming techniques in physics engine applications. With the advent of multi-core processors in contemporary computing environments, optimizing physics simulations through parallel programming has become increasingly feasible. A conventional blob collision physics engine serves as the benchmark for evaluation, and its performance is juxtaposed against a parallel-programmed variant. Experimental findings indicate a significant reduction in computational time required for collision detection and response when parallel processing is employed. This efficiency gain is particularly pronounced in scenarios involving a large number of blobs, showcasing the scalability advantages of parallelization. Moreover, parallel programming facilitates optimal harnessing of multi-core processor capabilities, thereby enhancing the overall efficiency and performance of the physics engine in question. This study not only substantiates the technical merits of applying parallel programming but also illuminates the practical benefits, including resource-efficient operation and quicker simulation times. Consequently, the research provides valuable insights for developers and engineers aiming to fully exploit the capabilities of modern computing hardware in physics-based simulations and applications.

Acceleration of Hyperspectral Skin Cancer Image Classification through Parallel Machine-Learning Methods.

Hyperspectral imaging (HSI) has become a very compelling technique in different scientific areas; indeed, many researchers use it in the fields of remote sensing, agriculture, forensics, and medicine. In the latter, HSI plays a crucial role as a diagnostic support and for surgery guidance. However, the computational effort in elaborating hyperspectral data is not trivial. Furthermore, the demand for detecting diseases in a short time is undeniable. In this paper, we take up this challenge by parallelizing three machine-learning methods among those that are the most intensively used: Support Vector Machine (SVM), Random Forest (RF), and eXtreme Gradient Boosting (XGB) algorithms using the Compute Unified Device Architecture (CUDA) to accelerate the classification of hyperspectral skin cancer images. They all showed a good performance in HS image classification, in particular when the size of the dataset is limited, as demonstrated in the literature. We illustrate the parallelization techniques adopted for each approach, highlighting the suitability of Graphical Processing Units (GPUs) to this aim. Experimental results show that parallel SVM and XGB algorithms significantly improve the classification times in comparison with their serial counterparts.

Multivariate Stochastic Volatility Modeling via Integrated Nested Laplace Approximations: A Multifactor Extension

This study introduces a multivariate extension to the class of stochastic volatility models, employing integrated nested Laplace approximations (INLA) for estimation. Bayesian methods for estimating stochastic volatility models through Markov Chain Monte Carlo (MCMC) can become computationally burdensome or inefficient as the dataset size and problem complexity increase. Furthermore, issues related to chain convergence can also arise. In light of these challenges, this research aims to establish a computationally efficient approach for estimating multivariate stochastic volatility models. We propose a multifactor formulation estimated using the INLA methodology, enabling an approach that leverages sparse linear algebra and parallelization techniques. To evaluate the effectiveness of our proposed model, we conduct in-sample and out-of-sample empirical analyses of stock market index return series. Furthermore, we provide a comparative analysis with models estimated using MCMC, demonstrating the computational efficiency and goodness of fit improvements achieved with our approach.

A stochastic simulation-based chance-constrained programming model for optimizing watershed best management practices for nonpoint source pollution control under uncertainty

In this study, a stochastic simulation-based chance-constrained programming (SSCP) model is developed by integrating soil and water assessment tool (SWAT), generalized likelihood uncertainty estimation (GLUE), chance-constrained programming (CCP) and genetic algorithm (GA) for managing non-point source (NPS) pollution under uncertainty. The SSCP model improves upon conventional simulation-based optimization methods by effectively (i) reflecting the propagation of stochastic uncertainties from the SWAT to the optimization framework by the joint GLUE and CCP (GLUE-CCP) approach, and (ii) identifying the optimal implementation levels of best management practices (BMPs) under different constraint-violation risk levels. The SSCP is solved using a GA with parallel sampling technique after conversion into the corresponding deterministic version. The model is applied to determine the optimal types, sizes and locations of BMPs for NPS pollution control in the Hanjiang River basin, China. The results indicate that, under low-risk conditions, grassed waterways and filter strips should be implemented in the northern part of the basin and larger detention ponds of about 637.8 ∼ 850.4 ha should be installed in the northern and eastern parts of the basin; whereas under high-risk conditions, filter strips should be installed in the northeastern part of the basin and larger detention ponds of about 562.1 ∼ 749.5 ha should be installed in the northern and southwestern parts of the basin. Moreover, as the risk level decreases, stricter control of constraint violations leads to higher BMP implementation costs in order to mitigate the risk of excessive NPS pollution loads. Thus, the proposed model provides valuable insights into the selection of BMP placement schemes at various risk levels, facilitating the reduction of NPS pollution at the outlet of the Hanjiang River basin to acceptable levels while considering the tradeoff between system cost and risk.

Development of a GPU-based multisphere DE–FE coupling method for interaction simulations between an off-road tire and a gravel terrain

For decades, growing interest has been devoted to simulating the tire–soil interaction phenomena using discrete element/finite element (DE–FE) coupling methods. In contrast to the widely used approach that idealizes soils as spherical DEs, a newly developed multisphere DE–FE method has been demonstrated to be more accurate in reproducing such phenomena; however, it requires considerable computational cost for large-scale tire–soil interaction simulations. Therefore, this study aims to develop an efficient graphics processing unit (GPU)-based multisphere DE–FE method based on a CUDA FORTRAN programming environment. To this end, two-level grids are introduced for the parallelization of contact detection for targeted multisphere DEs and subspheres, which helps to find contact pairs for the targeted subspheres of a multisphere DE in an efficient manner. In contrast to existing GPU-based parallelization techniques, which typically use atomic operations to avoid memory access conflicts, this study presents a parallelization strategy to process a contact pair between bounding spheres using one GPU thread to efficiently handle with owner-member relationships between multisphere DEs and their subspheres. In the parallelization for force calculation, our method introduces an efficient technique that utilizes one GPU thread to process a contact pair between subspheres, or between subspheres and FE segments, to reduce the number of repeated calculations. Finally, the developed GPU-based method is applied to analyze the travelling behaviors of a complex-patterned off-road tire on a gravel terrain, and the effectiveness and computational efficiency of this method are validated via indoor soil-bin experimental outcomes as well as serial and parallel simulation results. It is found that a speedup of approximately 38 has been achieved with the aid of the developed method.

3D-DESS MRI with CAIPIRINHA two- and fourfold acceleration for quantitatively assessing knee cartilage morphology.

This study aimed to assess the diagnostic image quality and compare the knee cartilage segmentation results using a controlled aliasing in parallel imaging results in higher acceleration (CAIPIRINHA)-accelerated 3D-dual echo steady-state (DESS) research package sequence in the knee. A total of 64 subjects underwent both two- and fourfold CAIPIRINHA-accelerated 3D-DESS and DESS without parallel acceleration technique of the knee on a 3.0T system. Two musculoskeletal radiologists evaluated the images independently for image quality and diagnostic capability following randomization and anonymization. The consistency of automatic segmentation results between sequences was explored using an automatic knee cartilage segmentation research application. The descriptive statistics and inter-observer and inter-method concordance of various acceleration sequences were investigated. P values < .05 were considered significant. For image quality evaluation, the image signal-to-noise ratio and contrast-to-noise ratio decreased with the decrease in scanning time. However, it is accompanied by the reduction of artifacts. Using 3D-DESS without parallel acceleration technique as the standard for cartilage grading diagnosisand the diagnostic agreement of two- and fourfold CAIPIRINHA-accelerated 3D-DESS was good, kappa value was 0.860 (P < .001) and 0.804 (p < 0.001), respectively. Regarding cartilage defects, the sensitivity and specificity of the twofold acceleration 3D-CAIPIRINHA-DESS were 95.56% and 97.70%, and the fourfold CAIPIRINHA-accelerated 3D-DESS were 91.49% and 97.65%, respectively. The intraclass correlation coefficients of various sequences in cartilage segmentation were almost all greater than 0.9. The CAIPIRINHA-accelerated 3D-DESS sequence maintained comparable diagnostic and segmentations performance of knee cartilage after a 60% scan time reduction.

Parallel programming: Driving the computational surge in AI

With the meteoric ascent and ongoing advancements in the realm of Artificial Intelligence (AI), there's an escalating demand for potent computational capabilities. Meeting the exacting computational demands of burgeoning AI technologies has become a pivotal topic in contemporary research. This study is anchored on the theme: "Impact of Parallel Programming on AI." The objective is to delve into the potentialities and value that parallel programming technology might introduce to the AI sector. To deeply probe the integration of parallel programming in AI, we've employed a meticulous research methodology. This process disentangles and scrutinizes the prevalent parallel programming techniques and their tangible applications within the AI sphere. Such a method offers a nuanced grasp of the fusion of parallel programming with AI and the distinct advantages that ensue. Findings reveal a substantial enhancement in the efficacy of AI models when they leverage parallel programming techniques. This is especially salient in scenarios involving large-scale data training and intricate model architectures. Most notably, parallel programming dramatically slashes AI training durations. This swift training paves the way for rapid iterations and fine-tuning. In conclusion, parallel programming emerges as a game-changer in the AI domain. It doesn't merely amplify AI performance but also lays a robust technical bedrock for AI's sustained and swift evolution. This opens up a vista of novel prospects and avenues for forthcoming exploration and application in AI.