Static Memory Research Articles

The association between memory, COVID-19 testing, and COVID-19 incidence in middle-aged and older adults: a prospective analysis of the CLSA

ABSTRACT We investigated the association between pre-COVID-19 memory function and (a) receipt of a COVID-19 test and (b) incidence of COVID-19 using the COVID-19 Questionnaire Study (CQS) of the Canadian Longitudinal Study on Aging (CLSA). The CQS included 28,565 middle-aged and older adults. We regressed receipt of a COVID-19 test on participants’ immediate and delayed recall memory scores and re-ran the regression models with COVID-19 incidence as the outcome. All regression models were adjusted for sociodemographic, lifestyle, and health covariates. In the analytical sample (n = 21,930), higher delayed recall memory (better memory) was significantly associated with lower COVID-19 incidence. However, this association was not significant for immediate recall memory. Immediate and delayed recall memory were not associated with receipt of a COVID-19 test. Health policymakers and practitioners may viewmemory status as a potential risk for COVID-19. Memory status may not be a barrier to COVID-19 testing.

A new [formula omitted] control method of singular time-delay systems based on a variable memory state feedback

This paper is concerned with the H∞ control problem of singular time-delay systems. A variable memory state feedback is established with state decomposition. It is verified that this feedback including the traditional memory and memoryless state feedbacks as special cases, and thus it is more general than these existing feedback strategies Based on the proposed feedback, a new H∞ control method with better performance level is derived. Moreover, the yielded computational complexity is lower. The superiority is confirmed by two given examples.

Far-gate synaptic transistors utilizing ion-charge dual-transfer mechanism for neurotransmitter-multiplexing temporal coding

The ability of artificial synapses to replicate multiplexed-transmission is a significant advancement in emulating complex brain activities. However, it generally required more stringent material requirements of intrinsic-ambipolarity and more complex structures of P/N dual-channel. Here, we proposed a far-gate synaptic transistor (FGST) just using a single-channel composed of a common unipolar semiconductor to emulate the cooperation and competition between two excitatory neurotransmitters. FGST exhibits a unique ion-charge dual-transfer mechanism, enabling distinct behavioral regulation modes with switchable synaptic plasticity: ion-dominant potentiation-depression with short-term plasticity and hole-dominant potentiation with enhanced memory. Moreover, FGST with dual-excitatory enhancement can be used for temporal contrast encoding, dividing currents into multiple memory states based on a fixed threshold; by comparing the variations in postsynaptic currents with different thresholds, it offers a method for further expanding the number of temporal states in the device. This work is a step toward constructing multifunctional artificial intelligent systems.

Acute stress impairs intentional memory suppression through aberrant prefrontal cortex activation in high trait ruminators

ObjectiveResearch shows that the effect of acute stress on intentional memory suppression could be modulated by individual differences in psychological traits. However, whether acute stress distinctly affects intentional memory suppression in high trait ruminators, a high at-risk group of stress-related disorders, and the neural correlations, remains unclear. Method55 healthy college students were divided into high and low trait ruminators (HTR and LTR), Following stress manipulation, a Think/No Think task assessed the memory suppression performance. Functional near-infrared spectroscopy was applied to explore the neural correlates. Psychophysiological interaction analyses were used to assess how the functional connectivity between a seed region and another brain region was modulated by tasks during memory suppression, further mediating memory suppression performance and state rumination. ResultsThe HTR exhibited poorer memory suppression performance than the LTR under the stress condition. Aberrant activation patterns and task-modulated functional connectivity in the dorsal prefrontal cortex (DLPFC) and superior temporal gyrus (STG) were observed only in the HTR during memory suppression under the stress condition. The effect of memory suppression performance on the state rumination of individuals was significantly mediated by the task-modulated functional connectivity between the DLPFC and STG. ConclusionsThe findings could provide insights for prevention or early intervention in the development of stress-related disorders in HTR.

The Effects of Different Exercise Types on Cognitive and Physical Functions in Dementia Patients: A Randomized Comparative Study

Purpose: The physical and cognitive effects of aerobic exercise on dementia have been extensively studied. Further investigation of other types of exercise with different physiological effects is still needed. This study aimed to determine cognitive and physical effects of 6-week aerobic, balance, and combined (aerobic-balance) exercise programs on dementia. Materials and Methods: A total of 31 mild to moderate dementia patients aged 65-90 years were divided into three exercise groups. Before and after the 6-week exercise program, mental rotation, spatial orientation, visual memory, and mental status were assessed for cognitive functions, while fall risk, reaction time, lower limb strength, and frailty were assessed for physical functions. Comprehensive cognitive and physical assessments were performed to provide a holistic approach to dementia. Results: When post-exercise values were compared with pre-exercise values, only frailty decreased significantly in the aerobic exercise group (p=0.017). After exercise program in balance and combined exercise groups, mental rotation (p=0.005, p=0.032), spatial orientation (p=0.020, p=0.035), mental status (p=0.007, p=0.014), and lower extremity strength (p=0. 010, p=0.005) increased significantly, while fall risk (p=0.005, p=0.005), reaction time (p=0.028, p=0.016), and frailty (p=0.020, p=0.009) decreased significantly. Moreover, in contrast to combined and aerobic exercise, improvement in visual memory was also observed in the balance exercise group (p=0.016). Conclusions: These findings suggest that balance and combined exercises may have broader effects on dementia than aerobic exercise. It emphasizes the importance of designing exercise programs for dementia patients, considering the cognitive and physical deficits of the patients, and creating a multidimensional treatment approach.

Exploring Flip Flop memories and beyond: training Recurrent Neural Networks with key insights.

Training neural networks to perform different tasks is relevant across various disciplines. In particular, Recurrent Neural Networks (RNNs) are of great interest in Computational Neuroscience. Open-source frameworks dedicated to Machine Learning, such as Tensorflow and Keras have produced significant changes in the development of technologies that we currently use. This work contributes by comprehensively investigating and describing the application of RNNs for temporal processing through a study of a 3-bit Flip Flop memory implementation. We delve into the entire modeling process, encompassing equations, task parametrization, and software development. The obtained networks are meticulously analyzed to elucidate dynamics, aided by an array of visualization and analysis tools. Moreover, the provided code is versatile enough to facilitate the modeling of diverse tasks and systems. Furthermore, we present how memory states can be efficiently stored in the vertices of a cube in the dimensionally reduced space, supplementing previous results with a distinct approach.

Maintenance of color memoranda in activity-quiescent working memory states: Evidence from impulse perturbation

In the present study, we used an impulse perturbation method to probe working memory maintenance of colors in neurally active and activity-quiescent states, focusing on a set of pre-registered analyses. We analyzed the electroencephalograph (EEG) data of 30 participants who completed a delayed match-to-sample working memory task, in which one of the two items that were presented was retro-cued as task relevant. The analyses revealed that both cued and uncued colors were decodable from impulse-evoked activity, the latter in contrast to previous reports of working memory for orientation gratings. Decoding of colors from oscillations in the alpha band showed that cued items could be decoded therein whereas uncued items could not. Overall, the outcomes suggest that subtle differences exist between the representation of colors, and that of stimuli with spatial properties, but the present results also demonstrate that regardless of their specific neural state, both are accessible through visual impulse perturbation.

Abstract 6424: A patient centric full spectrum flow cytometry assay for comprehensive phenotyping of CAR T cells in clinical trials

Abstract Purpose While there is a strong desire to fully understand next generation CAR-T product characteristics in clinical trials, limited amounts of blood specimens are available as the enrolled patients are very sick and lymphodepleted. To address this high unmet need, we developed a single tube 20-color full spectrum flow cytometry assay for an in-depth characterization of CAR-T cells that support fulfilment of five key clinical trial endpoints, namely, manufacturing success (fitness and memory state), pharmacokinetics, response or resistance (ICIs), efficacy (gross tumor burden) and safety (B cell regeneration) for use across a broad range of CAR-T therapies and clinical indications. Study Design: The single tube 20-parameter full spectrum flow cytometry (FC) assay was designed to include CD45, HLA-DR, CD3, CD4, CD8, CD56 as backbone markers, CD45RA, CCR7, CD95, CD25 and CD127 antibodies allowed for the identification of all major subsets of CD4 and CD8 T cells. CXCR3, CD38, PD1, LAG3, and TIM3 were used for identification of activation/exhaustion phenotypes, while CD20 and HLA-DR were useful for identification of B cells and monocytes respectively. Additionally, we left two parameters (PE and AF647) open for customized CAR-T detection reagents or other customized markers. This assay was developed using Cytek® Northern Lights™ and frozen PBMCs. Results This robust 20-color full spectrum FC assay allowed a thorough analysis of the major T cell populations cells. FMO (Fluorescence-Minus-One) control studies confirmed the specificity of the antibodies and of the panel in identifying Naïve, EMRA, CM, EM, Th17, T stem cell memory and regulatory T cells with minimum background on all parameters. This method showed to be highly reproducible and can be adapted for use with many CAR-T detection reagents. Conclusion In summary, we developed a robust 20-parameter multiparameter full spectrum flow cytometry assay focused on different populations of T cells in a single tube. The inclusion of drop ins for prospective CAR-T cell detection reagents allows for easy integration of future advances in CAR-T cell research. This universal panel shows great potential for usage across CAR-T clinical trials, as it serves as a valuable tool understand product fitness, PK, biomarker monitoring and disease burden simultaneously. Citation Format: Keegan Vaughan, Selvam Muthusamy, Coral Zhao, Austin Carbajal, Charlie Fu, Naveen Dakappagari, Zeni Alfonso. A patient centric full spectrum flow cytometry assay for comprehensive phenotyping of CAR T cells in clinical trials [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 6424.

Abstract 4020: Donor-dependent anti-tumoral efficacy of human CD19 CAR T cells in a leukemic xenograft mouse model

Abstract The recent advances in cellular immunotherapies have revolutionized the treatment options for hematological malignancies. Among others, genetically engineered T cells to express chimeric antigen receptors targeting CD19 have shown clinical success in patients with B cell malignancies. However, depending on the study design only one to two thirds of the patients experience a complete response upon CD19 CAR T cells treatment. Antigen escape, immune suppressive microenvironment, CAR T cell dysfunction as well as lack of CAR T cell persistence have been discussed to be reasons for the lack of sustained therapy response. Well established models for the preclinical evaluation of cellular therapies are needed for the evaluation of next generation CAR T cells as well as novel combination therapies. Here, we describe a xenograft animal model of human acute lymphoblastic leukemia (NALM-6 cells) with a suboptimal treatment response to human CD19 CAR T cells from three different donors. First, the activation status and memory phenotype of the CAR T cells from the different donors was characterized by flow cytometry and the anti-tumoral efficacy of the CD19 CAR T cells was evaluated in an in vitro co-culture killing assay using luciferase expressing NALM-6 cells. Next, CD19 CAR T cells from three donors were injected in NALM-6_luc tumor bearing animals. The tumor growth was monitored by bioluminescence imaging and the phenotype of the transferred CAR T cells was checked by flow cytometry. The frequency of memory T cells differed between the donors whereas the T cells from all donors exhibited all low expression of activation markers. In vitro, all three donors showed high and comparable anti-tumoral killing efficacy independent of their memory status. In vivo, CD19 CAR T cells from all three donors delayed the tumor growth significantly although none of the animals had a complete remission. In line with the individual memory status of the CAR T cells prior to infusion, there were slight differences in the kinetics of the anti-tumoral response. The flow cytometric analysis revealed that the transferred CAR T cells were almost absent in the peripheral blood and spleen whereas for individual animals tumor cells and highly activated CAR T cells were detectable in the bone marrow of the animals. In summary, the NALM-6_luc xenograft model is a value tool for evaluating next generation cellular therapies and novel combination strategies to overcome the current limitations of cellular therapies. Citation Format: Philipp Metzger, Carla N. Castro, Jonas F. Hummel, Maria Silvia Roman Azcona, Dinu Antony, Claudio Mussolino, Daniel Zurr, Swetlana Adamsky, Holger Weber. Donor-dependent anti-tumoral efficacy of human CD19 CAR T cells in a leukemic xenograft mouse model [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 4020.

Abstract 45: Driving AMPK signaling in CD4+ CART cells enhances metabolic capacity, memory formation, and anti-leukemia activity

Abstract Background: While Chimeric Antigen Receptor (CAR)T cell therapy has seen great success in pediatric and young adult acute lymphoblastic leukemia, up to 50% of patients will relapse following treatment. Increased memory formation and mitochondrial capacity, particularly in the CD4+ subset, have been linked to improved CART cell function and enhanced in vivo persistence, leading to long-term relapse-free survival. AMP-activated protein kinase (AMPK) is a cellular energy sensor responsible for promoting metabolic efficiency and T cell memory formation. We have previously demonstrated AMPK activity can be increased in human T cells via overexpression of the AMPKγ2 domain. We hypothesized that overexpression of AMPKγ2 in CART cells would enhance their metabolic capacity and improve their anti-leukemia function. Methods: Human CART cells bearing a CD19-CAR utilizing a 41BB costimulatory domain were generated from healthy human donors via lentiviral transduction and co-transduced with AMPKγ2 (AMPK) or an empty vector (EV) control. Flow sorting for the co-transduced populations yielded >95% CD4+ CART cells. CARTs were expanded in IL-2 for approximately 2 weeks, followed by overnight stimulation and subsequent metabolic analysis using the Seahorse Metabolic Analyzer. Memory status was assessed by CD62L and CCR7 co-expression. In vitro CART function was assessed using Zs-green+ NALM6 targets, with loss of green fluorescence over time measured on the Incucyte analyzer. Where indicated, tumor conditioned media was harvested from 72-hour NALM6 cultures. For in vivo experiments, immunodeficient NOD-scid IL2Rγnull (NSG) mice were injected with luciferase+ NALM6 targets, followed one week later by AMPK- or EV-CARTs. Leukemia burden, via IVIS imaging, and survival were followed over time. Results: CD4+ CARTs co-transduced with AMPKγ2 (AMPK-CARTs) increased oxygen consumption, maximal respiration, and spare respiratory capacity following overnight stimulation by ~20% compared to CARTs co-transduced with the EV (EV-CARTs) (p<0.05). Following expansion, CD62L+CCR7+ central memory cells were 4.5-fold more abundant in the AMPK-CART group (2.4-6.2-fold, p<0.01). In vitro, while AMPK-CARTs demonstrated similar cytotoxicity against NALM6 targets as EV-CARTs in standard media, they mediated faster killing of NALM6 targets when challenged in tumor conditioned media (p<0.001). In vivo, recipients of AMPK-CARTs showed a consistent decrease in tumor burden as measured by median radiance values, translating to a significant survival advantage (p<0.05). Conclusions: Increasing AMPK signaling in CD4+ CARTs via overexpression of AMPKγ2 improved metabolic profile, increased memory generation, and enhanced toxicity against leukemia targets both in vitro and in vivo. Together, these features make elevating AMPK signaling an attractive method to increase the efficacy of CART therapy. Citation Format: Erica Braverman, Herbert Schuler, Mengtao Qin, Craig Byersdorfer. Driving AMPK signaling in CD4+ CART cells enhances metabolic capacity, memory formation, and anti-leukemia activity [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 45.

An Integrated Lagrangian Modeling Method for Mechanical Systems with Memory Elements

Mechanical memory elements cannot be accurately modeled using the Lagrangian method in the classical sense, since these elements are nonconservative in the plane of their non-constitutive relationships, and the system differential equations are not self-adjoint and therefore do not allow a Lagrangian formulation. To overcome this problem, the integrated Lagrangian modeling method is introduced, in which the associated conventional energies in the system are replaced by the corresponding memory state functions of the memory elements. An example, a vehicle shimmy system equipped with fluid mem-inerters, is presented to verify the improvement of modeling accuracy of mechanical systems with memory elements via the integrated Lagrangian method. The simulation results show that under pulse and random excitation, using the Lagrangian method to model the system, the values of system response indicators exhibit significant errors ranging from 5.17% to 24.54% compared with the values obtained by the integrated Lagrangian method, namely, the accurate values. In addition, the influencing factors of the error and are discussed and the fractional-order memory elements and their modeling are also briefly generalized.

Exploring the Pedagogical Palette: Assessing Instructional Strategies and Proficiency among Higher Education Educators in the Philippines

This study evaluates the utilization of various instructional pedagogies and the level of competence among teachers at Don Mariano Marcos Memorial State University (MSU) and the Philippine College of Ministry (PCM) despite the challenges posed by the pandemic. Through a descriptive survey method, data was gathered from 43 teachers across both institutions. Results indicate no significant difference in the extent of utilization or competence in instructional pedagogies between DMMMSU and PCM. Additionally, a significant relationship was found between the extent of utilization and the level of competence among teachers in both institutions. These findings underscore the importance of ongoing training and collaboration between teachers and administrators to enhance pedagogical skills, particularly in the context of the pandemic. Recommendations include encouraging participation in training seminars, strategic planning for adapting to the "New Normal," and providing similar opportunities for students to enhance their understanding of instructional pedagogies.

IGZO charge trap flash device for reconfigurable logic functions

We fabricated an indium gallium zinc oxide (IGZO) charge trap flash (CTF) device for logic-in-memory (LIM) applications. Initially, the nonvolatile memory characteristics of the IGZO CTF device were investigated under charge trapping and detrapping states in 104 s retention tests. Next, we constructed a common-source amplifier circuit containing the IGZO CTF device and demonstrated various input–output signal relationships by modulating the memory state of the device. Finally, we used interconnected IGZO CTF devices to demonstrate reconfigurable logic functions. Using series- and parallel-connected IGZO CTF devices, we developed 2-input NAND and 2-input NOR gates, respectively. Our experimental results showed that the IGZO CTF device is a promising future memory device and a tool for LIM technology.

Пам’ять у лінгвістичному і соціокультурному вимірах

The article explores the linguistic characteristics of the concept of memory as one of the key cultural concepts among other meaningful ones — person, life, language, word. The starting point was the position that memory is a psychological category that is verbalized in different ways by linguistic means, models synonymous series, mental image forms. It is noted that the word-concept memory correlates with the words consciousness, memory, imagination, mind, knowledge through the prism of individual and collective experience of the past, and acts as a tool for evaluating human existence. In a comparative aspect, the semantics of the Ukrainian verb to remember and the Czech zapomnět, as well as the related nouns nezabudka and pomněnka, are considered. Verbs, having at first glance a similar grammatical structure, are characterized by antonymic meanings in modern literary languages. The factors that led the original semantic feature of the root pomn- (memor- [пам’ят-]) — «to remember» — to develop to the opposite meaning in the Czech verb zapomnět «to forget / not to remember something» are analyzed. It was noted that under the influence of other Slavic languages, the parallel use in the XVI–XVII c. in the Ukrainian language, two words — «пам’ятати / помнити» — led to the fact that the word to remember displaced the word to remember from usage. The semantics of ʽmemory’ in the interpretation of figurative means — metaphors and phraseology, which show signs of high semantic stability, reproducibility in the language from memory, and thus have the status of collective memories. The evidence presented in the article shows that the conceptual field of memory is structured by concepts that are relevant both for the individual and for the whole human community in a particular time dimension and have a high axiological potential.

Seeking for associations of I/D polymorphism of ACE gene with self-reported work performance, memory, and cardiovascular health in a professional leadership community

Introduction. Maintaining labor efficiency, working under conditions of stress and uncertainty, high level of requirements to the functional state of workers in various professions actualize the occupational health study of work performance. Methodological applications of predictive medicine and psychogenetics make it possible to identify specific hereditary (biological) predispositions of individuals, to carry out professional orientation of a person, to provide humanistic recommendations on occupation of a certain type of professional activity.
 The study aims to identify a correlation between ACE genotypes and work performance capability in individuals from a professional leadership community.
 Materials and methods. Individuals from a professional leadership community, after dividing into two cohorts consisting of entrepreneurs and non-entrepreneurs, were interviewed and genotyped for evidence of genotype-by-insertion/deletion (I/D) polymorphism associations of the angiotensin I converting enzyme (ACE) gene and health status. Answers to 4 questions characterized the state of work performance and memory, and another 4 questions characterized the cardiovascular system. The mean age was 37±1.7 years for entrepreneurs and 34.23±0.66 years for non-entrepreneurs. There were 44 females and 129 males among them.
 Results. Analysis of genotyping of 173 people (25 entrepreneurs and 148 non-entrepreneurs) revealed a significant deviation of actual frequencies of I and D alleles from theoretically expected values in the whole group of managers and between non-entrepreneurs (p=0.95, χ2>3.8). In all groups, there was a marked increase in the frequency of the I/I genotype at the expense of a decrease in the frequency of the D/D genotype compared to the literature data. A significant difference between D/D and I/D genotypes in work performance was found in entrepreneurs (p=0.9817), which coincides with a significant difference between their representatives by age (p>0.99). Contrary to expectations, representatives of the D/D genotype, at least entrepreneurs, demonstrated the most optimal level of functioning of the autonomic nervous system. No associations were found between the genotype for the ACE gene (rs4646994) and work performance and memory.
 Conclusion. In the total sample of leaders and among the subgroup of non-entrepreneurs, a strong and reliable change in the frequencies of alleles I and D was found compared to the theoretically expected (frequencies). There was a significant increase in the frequency of I/I genotype in all groups. The significant difference in the rates between the D/D and I/D groups coincides with a significant difference between their representatives in terms of age. According to the results of test answers, representatives of entrepreneurs of D/D genotype had the most optimal level of vegetative nervous system functioning.

Collective neural network behavior in a dynamically driven disordered system of superconducting loops

Collective properties of complex systems composed of many interacting components such as neurons in our brain can be modeled by artificial networks based on disordered systems. We show that a disordered neural network of superconducting loops with Josephson junctions can exhibit computational properties like categorization and associative memory in the time evolution of its state in response to information from external excitations. Superconducting loops can trap multiples of fluxons in many discrete memory configurations defined by the local free energy minima in the configuration space of all possible states. A memory state can be updated by exciting the Josephson junctions to fire or allow the movement of fluxons through the network as the current through them surpasses their critical current thresholds. Simulations performed with a lumped element circuit model of a 4-loop network show that information written through excitations is translated into stable states of trapped flux and their time evolution. Experimental implementation on a high-Tc superconductor YBCO-based 4-loop network shows dynamically stable flux flow in each pathway characterized by the correlations between junction firing statistics. Neural network behavior is observed as energy barriers separating state categories in simulations in response to multiple excitations, and experimentally as junction responses characterizing different flux flow patterns in the network. The state categories that produce these patterns have different temporal stabilities relative to each other and the excitations. This provides strong evidence for time-dependent (short-to-long-term) memories, that are dependent on the geometrical and junction parameters of the loops, as described with a network model.

The Nonvolatile Memory and Neuromorphic Simulation of ReS2/h-BN/Graphene Floating Gate Devices Under Photoelectrical Hybrid Modulations.

The floating gate devices, as a kind of nonvolatile memory, obtain great application potential in logic-in-memory chips. The 2D materials have been greatly studied due to atomically flat surfaces, higher carrier mobility, and excellent photoelectrical response. The 2D ReS2 flake is an excellent candidate for channel materials due to thickness-independent direct bandgap and outstanding optoelectronic response. In this paper, the floating gate devices are prepared with the ReS2/h-BN/Gr heterojunction. It obtains superior nonvolatile electrical memory characteristics, including a higher memory window ratio (81.82%), tiny writing/erasing voltage (±8V/2ms), long retention (>1000s), and stable endurance (>1000 times) as well as multiple memory states. Meanwhile, electrical writing and optical erasing are achieved by applying electrical and optical pulses, and multilevel storage can easily be achieved by regulating light pulse parameters. Finally, due to the ideal long-time potentiation/depression synaptic weights regulated by light pulses and electrical pulses, the convolutional neural network (CNN) constructed by ReS2/h-BN/Gr floating gate devices can achieve image recognition with an accuracy of up to 98.15% for MNIST dataset and 91.24% for Fashion-MNIST dataset. The research work adds a powerful option for 2D materials floating gate devices to apply to logic-in-memory chips and neuromorphic computing.

Oxide Semiconductor Memristor‐Based Optoelectronic Synaptic Devices With Quaternary Memory Storage

AbstractA pioneering integration of oxide semiconductor memristors with optoelectronic features is presented, surpassing binary limitations to realize multi‐valued synaptic operations. Through Pt/Ga2O3/Pt memristors, their structural and electronic attributes via atomic force microscopy, X‐ray diffraction, and X‐ray photoelectron spectroscopy are explored. Demonstrating unipolar resistance switching with remarkable endurance and retention, the devices exhibit intricate light‐resistance correlations, yielding substantial photoelectric effects in distinct resistance states. Investigating synaptic behaviors, potentiation, and depression akin to biological synapses are unveiled, facilitating learning and memory processes. The standout achievement lies in attaining quaternary memory storage within a single device. Empirical data and simulations validate this concept, showcasing the potential for encoding and sustaining multiple memory states. This innovation heralds transformative possibilities, emphasizing oxide semiconductor memristors as a gateway to quaternary memory storage and enhanced synaptic functions. In essence, this work pioneers optoelectronic synaptic devices with expanded memory capabilities.

A static memory method for modelling time-fractional power law absorption

The attenuation of ultrasound propagation through tissue is known to follow a frequency power-law in the time domain. This can be modelled as a loss operator in the equation of state within the Euler equations that takes the form of a fraction time derivative operator. This can be treated through a second order accurate transfer between the fractional time derivative and a fractional Laplacian spacial operator in order to avoid storage of the full time history. This is used for example by the k-Wave toolbox. As an alternate finite history methods have been suggested. Building on recent work, here, we re-write the time fractional derivative as a finite sum using a recursion relation. This allows the time fractional derivative to be directly computed with a static memory requirement. For homogeneous media, the advantage of this over the fractional Laplacian methods may not initially be obvious with a base higher memory requirement and computational cost with only a small increase in accuracy. However, upon introducing a heterogeneous medium with regions of different power law attenuation; the increased computational cost of the time fractional method is contained exclusively within the pre-computation, while the increased cost for the fractional Laplacian is applied to each time step.

Live Migration of Video Analytics Applications in Edge Computing

In order to schedule resources efficiently or maintain applications' continuity for mobile customers, edge platforms often need to adaptively migrate the applications on them. However, our measurement shows that existing migration solutions cannot solve the issue of migrating video analytics applications in edge computing because the memory states of video analytics applications have different characteristics from other applications. We conduct a breakdown analysis of the memory states of video analytics applications, and propose to treat three types of states separately with three different techniques, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">i.e.</i> , warm-up, sync, and replay, to minimize the negative influence of migrations on application performance. Based on this idea, we implement a prototype system in which two new components, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">i.e.</i> , <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">state store</i> and <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">sidecar</i> , are designed to achieve near-transparent live migration with minimal application code modifications. Evaluation experiments demonstrate that the time of application interruption caused by migrating a video analytics application with our solution is less than 405ms, and our solution does not consume much resources.