Temporal Parallelism Research Articles

Rodent models for oral microbiome research: considerations and challenges- a mini review.

Rodent models have been commonly employed in oral microbiota research to investigate the relationship between bacteria and oral disease. Nevertheless, to apply the knowledge acquired from studies conducted on rodents to a human context, it is crucial to consider the significant spatial and temporal parallels and differences between the oral microbiota of mice and humans. Initially, we outline the comparative physiology and microbiology of the oral cavity of rodents and humans. Additionally, we highlight the strong correlation between the oral microbiome of rodents and genetic makeup, which is influenced by factors including vendor, husbandry practices, and environmental conditions. All of these factors potentially impact the replicability of studies on rodent microbiota and the resulting conclusions. Next, we direct our attention toward the diversity in the microbiome within mice models of disease and highlight the diversity that may potentially affect the characteristics of diseases and, in turn, alter the ability to replicate research findings and apply them to real-world situations. Furthermore, we explore the practicality of oral microbial models for complex oral microbial diseases in future investigations by examining the concept of gnotobiotic and germ-free mouse models. Finally, we stress the importance of investigating suitable techniques for characterizing and managing genetically modified organisms. Future research should consider these aspects to improve oral microbiome research's translational potential.

Performance evaluation of Word2vec accelerators exploiting spatial and temporal parallelism on DDR/HBM-based FPGAs

Word embedding is a technique for representing words as vectors in a way that captures their semantic and syntactic relationships. The processing time of one of the most popular word embedding technique Word2vec is very large due to the huge data size. We evaluate the performance of a power-efficient FPGA-based accelerator designed using OpenCL. We achieved up to 18.7 times speed-up compared to single-core CPU implementation with the same accuracy. The proposed accelerator consumes less than 83 W of power and it is the most power-efficient one compared to many top-end CPU and GPU-based accelerators.

Reducing time and memory requirements in topology optimization of transient problems

AbstractIn topology optimization of transient problems, memory requirements and computational costs often become prohibitively large due to the backward‐in‐time adjoint equations. Common approaches such as the Checkpointing (CP) and Local‐in‐Time (LT) algorithms reduce memory requirements by dividing the temporal domain into intervals and by computing sensitivities on one interval at a time. The CP algorithm reduces memory by recomputing state solutions instead of storing them. This leads to a significant increase in computational cost. The LT algorithm introduces approximations in the adjoint solution to reduce memory requirements and leads to a minimal increase in computational effort. However, we show that convergence can be hampered using the LT algorithm due to errors in approximate adjoints. To reduce memory and/or computational time, we present two novel algorithms. The hybrid Checkpointing/Local‐in‐Time (CP/LT) algorithm improves the convergence behavior of the LT algorithm at the cost of an increased computational time but remains more efficient than the CP algorithm. The Parallel‐Local‐in‐Time (PLT) algorithm reduces the computational time through a temporal parallelization in which state and adjoint equations are solved simultaneously on multiple intervals. State and adjoint fields converge concurrently with the design. The effectiveness of each approach is illustrated with two‐dimensional density‐based topology optimization problems involving transient thermal or flow physics. Compared to the other discussed algorithms, we found a significant decrease in computational time for the PLT algorithm. Moreover, we show that under certain conditions, due to the use of approximations in the LT and PLT algorithms, they exhibit a bias toward designs with short characteristic times. Finally, based on the required memory reduction, computational cost, and convergence behavior of optimization problems, guidelines are provided for selecting the appropriate algorithms.

Об одной особенности английского исторического романа рубежа ХХ–ХХI веков: размышляя над статьей Джерома де Гроота

Critically reading the chapter of the famous British specialist in the field of theory and history of the English historical novel, Jerome de Groot, ‘Transgression and Experimentation: A Historical Novel’ in the multiauthored monograph British Fiction: 1980–2018 (Cambridge, 2019), the author of the article reflects on the phenomenon of the primacy of the historical novel in the canon of the contemporary British novel both in the postmodernist and in the post-postmodernist eras. Agreeing with de Groot that the historical novel of the designated period not only preserved the core of the classical genre paradigm, laid down by Sir Walter Scott, but boldly stepped into modernity, modernizing both meaningfully and artistically the eternal dialogue of the past and present, originally inherent in the genre, the author of the article emphasizes how well the chapter fits into the general concept of the monograph – to show the dynamics of British literature precisely through the prism of the dialectic of tradition and experiment. The author of the article agrees with his English colleague that it has been since the 1980s that English contemporary historical novel began to plunge even more thoroughly and philosophically (ontologically), without responding too much to postmodern historical and epistemological doubts, into the eternal questions of existence, the complexity of the dialogue between man and the world, the fate of man and humanity, etc. The author of the article also agrees with the idea of the English literary critic that metafictional historiographical approach and self-reflection (according to Linda Hutcheon) in the British historical novel did not become absolutely dominant phenomena; that the English historical novel of the 1980s and 1990s and onwards, having quickly ‘overcome’ postmodern delights, still retained the tradition of socio-psychological realistic reading of history, even experimenting with narrative, time structure, and historiographical concepts, which was especially characteristic of the female reading of the past (A. Carter, J. Winterson, A. S. Byatt, S. Waters). Simultaneously with the postmodernist exercises inspired in the field of historical narrative and temporal parallel (synonymous, antithetical, alternative) historical constructs, another line matured in the English historical novel, not related to the postmodern game, but related to the desire to reproduce the past in its volume, in the integrity of its cultural appearance, to construct the image of the past, to the ability to immerse the reader in the very existence of history. With the example of the novels by H. Mantel, R. Tremaine, D. Moggach, etc. the author of the article demonstrates this side of the dynamics of the English historical novel, not emphasized by de Groot.

Fast reconstruction of SMS bSSFP myocardial perfusion images using noise map estimation network (NoiseMapNet): a head-to-head comparison with parallel imaging and iterative reconstruction.

Simultaneous multi-slice (SMS) bSSFP imaging enables stress myocardial perfusion imaging with high spatial resolution and increased spatial coverage. Standard parallel imaging techniques (e.g., TGRAPPA) can be used for image reconstruction but result in high noise level. Alternatively, iterative reconstruction techniques based on temporal regularization (ITER) improve image quality but are associated with reduced temporal signal fidelity and long computation time limiting their online use. The aim is to develop an image reconstruction technique for SMS-bSSFP myocardial perfusion imaging combining parallel imaging and image-based denoising using a novel noise map estimation network (NoiseMapNet), which preserves both sharpness and temporal signal profiles and that has low computational cost. The proposed reconstruction of SMS images consists of a standard temporal parallel imaging reconstruction (TGRAPPA) with motion correction (MOCO) followed by image denoising using NoiseMapNet. NoiseMapNet is a deep learning network based on a 2D Unet architecture and aims to predict a noise map from an input noisy image, which is then subtracted from the noisy image to generate the denoised image. This approach was evaluated in 17 patients who underwent stress perfusion imaging using a SMS-bSSFP sequence. Images were reconstructed with (a) TGRAPPA with MOCO (thereafter referred to as TGRAPPA), (b) iterative reconstruction with integrated motion compensation (ITER), and (c) proposed NoiseMapNet-based reconstruction. Normalized mean squared error (NMSE) with respect to TGRAPPA, myocardial sharpness, image quality, perceived SNR (pSNR), and number of diagnostic segments were evaluated. NMSE of NoiseMapNet was lower than using ITER for both myocardium (0.045 ± 0.021 vs. 0.172 ± 0.041, p < 0.001) and left ventricular blood pool (0.025 ± 0.014 vs. 0.069 ± 0.020, p < 0.001). There were no significant differences between all methods for myocardial sharpness (p = 0.77) and number of diagnostic segments (p = 0.36). ITER led to higher image quality than NoiseMapNet/TGRAPPA (2.7 ± 0.4 vs. 1.8 ± 0.4/1.3 ± 0.6, p < 0.001) and higher pSNR than NoiseMapNet/TGRAPPA (3.0 ± 0.0 vs. 2.0 ± 0.0/1.3 ± 0.6, p < 0.001). Importantly, NoiseMapNet yielded higher pSNR (p < 0.001) and image quality (p < 0.008) than TGRAPPA. Computation time of NoiseMapNet was only 20s for one entire dataset. NoiseMapNet-based reconstruction enables fast SMS image reconstruction for stress myocardial perfusion imaging while preserving sharpness and temporal signal profiles.

Guiding gaze gestures on smartwatches: Introducing fireworks

Smartwatches enable interaction anytime and anywhere, with both digital and augmented physical objects. However, situations with busy hands can prevent user inputs. To address this limitation, we propose Fireworks, an innovative hands-free alternative that empowers smartwatch users to trigger commands effortlessly through intuitive gaze gestures by providing post-activation guidance. Fireworks allows command activation by guiding users to follow targets moving from the screen center to the edge, mimicking real life fireworks. We present the experimental design and evaluation of two Fireworks instances. The first design employs temporal parallelization, displaying few dynamic targets during microinteractions (e.g., snoozing a notification while cooking). The second design sequentially displays targets to support more commands (e.g., 20 commands), ideal for various scenarios other than microinteractions (e.g., turn on lights in a smart home). Results show that Fireworks’ single straight gestures enable faster and more accurate command selection compared to state-of-the-art baselines, namely Orbits and Stroke. Additionally, participants expressed a clear preference for Fireworks’ original visual guidance.

Parallel-in-time integration of the shallow water equations on the rotating sphere using Parareal and MGRIT

Despite the growing interest in parallel-in-time methods as an approach to accelerate numerical simulations in atmospheric modeling, improving their stability and convergence remains a substantial challenge for their application to operational models. In this work, we study the temporal parallelization of the shallow water equations on the rotating sphere combined with time-stepping schemes commonly used in atmospheric modeling due to their stability properties, namely an Eulerian implicit-explicit (IMEX) method and a semi-Lagrangian semi-implicit method (SL-SI-SETTLS). The main goal is to investigate the performance of parallel-in-time methods, namely Parareal and Multigrid Reduction in Time (MGRIT) when these well-established schemes are used on the coarse discretization levels and provide insights on how they can be improved for better performance. We begin by performing an analytical stability study of Parareal and MGRIT applied to a linearized ordinary differential equation depending on some temporal parallelization parameters, including the choice of a coarse scheme. Next, we perform numerical simulations of two standard tests in atmospheric modeling to evaluate the stability, convergence, and speedup provided by the parallel-in-time methods compared to a fine reference solution computed serially. We also conduct a detailed investigation on the influence of artificial viscosity and hyperviscosity approaches, applied on the coarse discretization levels, on the performance of the temporal parallelization. Both the analytical stability study and the numerical simulations indicate a poorer stability behavior when SL-SI-SETTLS is used on the coarse levels, compared to the IMEX scheme. With the IMEX scheme, a better trade-off between convergence, stability, and speedup compared to serial simulations can be obtained under proper parameters and artificial viscosity choices, opening the perspective of the potential competitiveness for realistic models.

Speech emotion recognition via multiple fusion under spatial–temporal parallel network

Speech, as a necessary way to express emotions, plays a vital role in human communication. With the continuous deepening of research on emotion recognition in human–computer interaction, speech emotion recognition (SER) has become an essential task to improve the human–computer interaction experience. When performing emotion feature extraction of speech, the method of cutting the speech spectrum will destroy the continuity of speech. Besides, the method of using the cascaded structure without cutting the speech spectrum cannot simultaneously extract speech spectrum information from both temporal and spatial domains. To this end, we propose a spatial–temporal parallel network for speech emotion recognition without cutting the speech spectrum. To further mix the temporal and spatial features, we design a novel fusion method (called multiple fusion) that combines the concatenate fusion and ensemble strategy. Finally, the experimental results on five datasets demonstrate that the proposed method outperforms state-of-the-art methods.

Time parallelism and Newton-adaptivity of the two-derivative deferred correction discontinuous Galerkin method

In this work, we consider a high-order discretization of compressible viscous flows allowing parallelization both in space and time.The discontinuous Galerkin spectral element method, which is well-suited for massively parallel simulations, is used for spatial discretization. The main novelty in this work is the additional demonstration of time-parallel capabilities within an implicit two-derivative timestepping procedure to further increase the parallel speedup. Temporal parallelism is made possible by a predictor-corrector-type time discretization that allows to split the associated workload onto multiple processors.We identify a homogeneous load balance with respect to the linear (GMRES) iterations on each processor as a key for parallel efficiency. To homogenize the load and to enable practical simulations, an adaptive strategy for Newton’s method is introduced. It is shown that the time-parallel method provides a parallel efficiency of approx. 60−70% on 4−7 computational partitions. Moreover, the capabilities of the novel method for the simulation of large-scale problems are illustrated with a mixed temporal and spatial parallelization on more than 1000 processors.

SASA: A Scalable and Automatic Stencil Acceleration Framework for Optimized Hybrid Spatial and Temporal Parallelism on HBM-based FPGAs

Stencil computation is one of the fundamental computing patterns in many application domains such as scientific computing and image processing. While there are promising studies that accelerate stencils on FPGAs, there lacks an automated acceleration framework to systematically explore both spatial and temporal parallelisms for iterative stencils that could be either computation-bound or memory-bound. In this article, we present SASA, a scalable and automatic stencil acceleration framework on modern HBM-based FPGAs. SASA takes the high-level stencil DSL and FPGA platform as inputs, automatically exploits the best spatial and temporal parallelism configuration based on our accurate analytical model, and generates the optimized FPGA design with the best parallelism configuration in TAPA high-level synthesis C++ as well as its corresponding host code. Compared to state-of-the-art automatic stencil acceleration framework SODA that only exploits temporal parallelism, SASA achieves an average speedup of 3.41× and up to 15.73× speedup on the HBM-based Xilinx Alveo U280 FPGA board for a wide range of stencil kernels.

Temporal Parallelization of Dynamic Programming and Linear Quadratic Control

This article proposes a general formulation for temporal parallelization of dynamic programming for optimal control problems. We derive the elements and associative operators to be able to use parallel scans to solve these problems with logarithmic time complexity rather than linear time complexity. We apply this methodology to problems with finite state and control spaces, linear quadratic tracking control problems, and to a class of nonlinear control problems. The computational benefits of the parallel methods are demonstrated via numerical simulations run on a graphics processing unit.

Analysis of the main trends of modern Ukrainian wartime painting and artistic parallels

The purpose of the article is to highlight the main stylistic features of modern Ukrainian fine arts of the war period in relation to the art of Ukraine of the twentieth century. The research methodology is based on the comparative analysis of the visual-plastic language of contemporary Ukrainian art, as well as the comparative-historical method, which allows us to analyze the artistic temporal parallels of the development of Ukrainian fine arts. The scientific novelty consists in the first attempt to create a holistic picture of the development of modern Ukrainian fine arts of the war period and to determine the main artistic trends in the formation. Conclusions. The war period for Ukrainian art became a milestone that separated the artistic trends of the early XXI century and stylistic features of 2022, which introduced new artistic priorities and requirements. As a result of the analysis, it is established that the main trends in the development of modern Ukrainian wartime painting can already be identified. Generally speaking, this is primarily a tendency to posterity, generalization, monumentality. Heroization of the images of the victorious pages of the past in Ukrainian history, as well as glorification of the image of Ukraine-woman, suffering, but unshakable in the fight against the enemy. The use of Ukrainian symbols in works of art - flag, coat of arms, Ukrainian ornament, etc. - acquires a new sound. An attempt was also made to structure works of modern painting in the following areas: features of graphics and posters; tendencies of folk art and neo-primitivism; signs of heroic and epic style, neorealism. Of course, this process of forming a new plastic trend in Ukrainian painting has only just begun; it develops, moves, pulsates, changes. It is happening here and now. However, despite the terrible catastrophe and its destructive action, which we are witnessing and the name of which is the war, today we can see how radical changes are taking place not only in the national context of Ukraine, but around the world. And the destiny of art is not only to reflect these global changes that we have witnessed, but also to lead society, consolidate, affirm the eternal values and ideals of humanity, national foundations and lead to victory.&#x0D; Key words: modern painting, poster, national style, neorealism, neo-primitivism, avant-garde, national symbols.

FPGA-Based Pedestrian Detection for Collision Prediction System.

Pedestrian detection (PD) systems capable of locating pedestrians over large distances and locating them faster are needed in Pedestrian Collision Prediction (PCP) systems to increase the decision-making distance. This paper proposes a performance-optimized FPGA implementation of a HOG-SVM-based PD system with support for image pyramids and detection windows of different sizes to locate near and far pedestrians. This work proposes a hardware architecture that can process one pixel per clock cycle by exploring data and temporal parallelism using techniques such as pipeline and spatial division of data between parallel processing units. The proposed architecture for the PD module was validated in FPGA and integrated with the stereo semi-global matching (SGM) module, also prototyped in FPGA. Processing two windows of different dimensions permitted a reduction in miss rate of at least 6% compared to a uniquely sized window detector. The performances achieved by the PD system and the PCP system in HD resolution were 100 and 66.2 frames per second (FPS), respectively. The performance improvement achieved by the PCP system with the addition of our PD module permitted an increase in decision-making distance of 3.3 m compared to a PCP system that processes at 30 FPS.

TGL

Many real world graphs contain time domain information. Temporal Graph Neural Networks capture temporal information as well as structural and contextual information in the generated dynamic node embeddings. Researchers have shown that these embeddings achieve state-of-the-art performance in many different tasks. In this work, we propose TGL, a unified framework for large-scale offline Temporal Graph Neural Network training where users can compose various Temporal Graph Neural Networks with simple configuration files. TGL comprises five main components, a temporal sampler, a mailbox, a node memory module, a memory updater, and a message passing engine. We design a Temporal-CSR data structure and a parallel sampler to efficiently sample temporal neighbors to form training mini-batches. We propose a novel random chunk scheduling technique that mitigates the problem of obsolete node memory when training with a large batch size. To address the limitations of current TGNNs only being evaluated on small-scale datasets, we introduce two large-scale real-world datasets with 0.2 and 1.3 billion temporal edges. We evaluate the performance of TGL on four small-scale datasets with a single GPU and the two large datasets with multiple GPUs for both link prediction and node classification tasks. We compare TGL with the open-sourced code of five methods and show that TGL achieves similar or better accuracy with an average of 13X speedup. Our temporal parallel sampler achieves an average of 173X speedup on a multi-core CPU compared with the baselines. On a 4-GPU machine, TGL can train one epoch of more than one billion temporal edges within 1-10 hours. To the best of our knowledge, this is the first work that proposes a general framework for large-scale Temporal Graph Neural Networks training on multiple GPUs.

Design and Implementation of True Parallelism Quad-Engine Cybersecurity Architecture on FPGA

Applications, such as Internet of Things, deal with huge amount of transmitted, processed and stored images that required a high computing capability. Therefore, there is a need a computing architecture that contribute in increasing the throughput by exploiting modern technologies in both spatial and temporal parallelisms. This paper conducts a parallel quad-engine cybersecurity architecture with new configuration to increase the throughput. using DE1-SoC and Neek FPGA boards and HDL. In this architecture, each engine operates with 600MHz maximum frequency. Each image is divided into four parts of equal size and each part processed by single engine concurrently to achieve spatial parallelism. Internally, engine is handling image’s part in temporal parallelism and deep pipelining abstraction applied in every engine by dividing it to sub modules to execute different tasks concurrently. All data processed in engines is encrypted via AES algorithm that implemented as a significant part of engine architecture. The obtained results increased the throughput by four times, with 153,600Mbps, that make this computing architecture efficient and suitable for fast applications such as IoT and cybersecurity level of processing.

Implementation Dual Parallelism Cybersecurity Architecture on FPGA

This paper presents an efficient parallelism architecture that uses a dual-computing engine architecture to better throughput using both spatial and temporal parallelism on FPGA technology. This architecture will enhance the performance in terms of operating frequency and throughput and reduces the power consumption that meets applications with huge data processing such as Internet of Things .in this design, two boards are used, "DE1_Soc and NEEK board" with Altera Quartus Prime 18 for synthesis and simulation. The proposed design architecture gives better resource usage and throughput through fewer hardware redundancies using a frequency of 600MHZ with 64 bits for each engine from the dual-engine. Furthermore, the proposed architecture implementation results show the reduction in the time delay by 40 % and achieves a throughput of 153.6 Gb/s

Герметизм Володимира Свідзінського (у проєкції від міста над Смотричем)

The poetry of podolianyn Volodymyr Svidzynskyi is not hermetic itself, but fi rstly intellectual as the research proved. As the political situation for Ukrainian art is not favourable, it contains the notion from «big silence» to «worldview un-pronounceability», inspired by Podillia and Smotrych views. It is characterized by symbolism and surrealism. V. Svidzynskyi’s lyrics is noticiable for motives of detachment and non-interference as amplitude dramatic prophecy. It is full of martyr feeling from fatality talent. All of the mentioned above according the researchers equal V. Svidzynskyi to the world discourse of tragic celebrities as Paul Celan, Rainer Maria Rilke, Federico Garcia Lorca, Borys Pasternak. That is lyrical impreza of the poet was founded on the intellectuality worldview and it considers as avant-garde refl ection of the epoch. It was in great demand during the 21st century and established appropriately at that time too.In the fi rst and second collections of poems, Volodymyr Svidzinskyi was no-ticeably inclined to symbolism, he became interested in the plasticity of images, he turned them into symbols through a metaphor.Volodymyr Svidzinskyi, in the days of the so-called third Ukrainian renais-sance (that is, the «shot revival»), rethought himself as a medium. Although this motif became popular in the days of the scientifi c and technological revolution, in the poetry of the sixties, in particular, in the poems of Lina Kostenko (es-pecially in the collection «River of Heraclitus» and «Madonna’s Crossroads»).In the last two collections we can more and more clearly trace the infl uence of surrealism. However, we can not say that the poet was fascinated by the paint-ings of his famous contemporary Salvador Dali. The authors were simply in the same temporal parallel, did not imitate each other, did not artifi cially use the sur-real elements popular at that time in Western Europe.

Measured Life: Making Live, the “Modern System of Science,” and the Animated Bodies of Frankenstein

Abstract This article considers Mary Shelley’s novel Frankenstein through what Sara Guyer calls “biopoetics,” hybridizing biopolitical and romantic reading strategies, and positing that romantic writing arises in temporal, theoretical, and political parallel with the movement of power from the reign of the sovereign to the realm of biopower. I focus on how Frankenstein imagines the flesh of Victor as animated and directed forward through biopower, by way of the novel’s juxtaposed medico-scientific and romantic discourse of life. Through close readings of the creation scene and Victor’s final breaths aboard Walton’s exploratory Arctic ship, I conclude that Frankenstein at last offers itself both as artifact and archaeology of modern power—or what Guyer calls “literature as a form of biopower.”

Highly-Parallel FPGA Accelerator for Simulated Quantum Annealing

Quantum annealing (QA) is a method to find the global optimum of a combinatorial optimization problem by using quantum fluctuations. Quantum annealers such as D-wave are efficient to solve small problems with less than 2048 variables. Simulated quantum annealing on digital computers allows us to solve large real-world problems. However, the processing time increases exponentially with the number of variables. This article proposes a highly-parallel accelerator for simulated quantum annealing exploiting spatial and temporal parallelism. The accelerator is implemented using “open computing language (OpenCL)” on FPGA. For 8,192 spin models, we achieve 145 times speed using 32 Trotters in one FPGA and 290 times speed-up using 64 Trotters in two FPGAs, compared to single-core CPU implementation.

FPGA-based Design of Multiple Models for Industry 4.0 Cyber Security

Industry 4.0 computing infrastructure depends on traditional processing platforms, and all these platforms perform well and achieving acceptable throughput. Two significant problems confront with these platforms are the delay in data processing and security hazard for potential threats. A design coding has been written using VHDL based on behaviour temporal parallelism for Industry 4.0 cybersecurity model. The models are divided into three namely, single, dual and quad models. Every model has three main components include UART, FIFO and DES algorithm for data encryption and decryption. Encryption is achieved more than one block at a time and over the number of designs to find the most suitable application for data processing. A DES algorithm based efficient implementing the design of 16-round multiple parallel models for low-cost, scalable, and robust encryption solution of a maximum clock frequency of 227MHz and it is capable of encrypting or decrypting data blocks at a rate of throughput 58,288Mbps.