SearcharxivSearch

arXiv subjects

Sangeeta Yadav

Publications and source records attributed to Sangeeta Yadav.

3 recordsLinked to original sources

Two-dimensional clay channels for tunable nanofluidic memristor

Dynamic reconfiguration of charge carriers in confined ion-channels under electrical stimulation produces memory effects, where the internal resistance depends on history of the electric field. Vermiculite nanofluidic devices harness this effect to store and process information within a single component. We report switching between distinct memory loops by tuning ion transport pathways, governed by asymmetrical device architecture and intrinsic surface-charge. Polarity-dependent memory switching between crossing-1 and crossing-2 loops is achieved solely by altering electrode configurations, without modifying electrolyte, channel surface chemistry or device structure: providing mechanistic insights into ionic memristors through a straightforward, experimental strategy. The memristive characteristics are demonstrated in both in-plane and out-of-plane channel configurations with channel lengths spanning from centimeters to micrometers length scales using re-stacked vermiculite membranes and further investigated for miniaturization with devices of nanometer scale channel lengths, fabricated via ultramicrotomy method. Furthermore, we demonstrate neuromorphic functionalities, including synaptic potentiation-depression and programmable memory retention, highlighting potential for bio-inspired computing systems. Cost-effective and scalable fabrication solution processed vermiculite membrane memristors pave the way for practical integration of nanofluidic memristors for neuromorphic computing applications.

cond-mat.mtrl-sci

AI-augmented stabilized finite element method

An artificial intelligence-augmented Streamline Upwind/Petrov-Galerkin finite element scheme (AiStab-FEM) is proposed for solving singularly perturbed partial differential equations. In particular, an artificial neural network framework is proposed to predict optimal values for the stabilization parameter. The neural network is trained by minimizing a physics-informed cost function, where the equation's mesh and physical parameters are used as input features. Further, the predicted stabilization parameter is normalized with the gradient of the Galerkin solution to treat the boundary/interior layer region adequately. The proposed approach suppresses the undershoots and overshoots in the stabilized finite element solution and outperforms the existing neural network-based partial differential equation solvers such as Physics-Informed Neural Networks and Variational Neural Networks.

math.AP

SP Async:Single Source Shortest Path in Asynchronous Mode on MPI

Finding single source shortest path is a very ubiquitous problem. But with the increasing size of large datasets in important application like social network data-mining, network topology determination-efficient parallelization of these techniques is needed to match the need of really large graphs. We present a new Inter node-bellman cum Intra node Dijkstra technique implemented in MPI to solve SSSP problem. We have used a triangle based edge pruning for idle processes, and two different techniques for termination detection. Within each node the algorithm works as Dijkstra and for outer communication it behaves as inter node bellman ford. First termination detection technique is based on the token ring and counter. Second is a heuristic based technique, in which the timeout is calculated from the number of inter-edges and number of partitions. In this project asynchronous mode of message passing is used.

cs.DC