More Like this

1. Imaging of bi-anisotropic periodic structures from electromagnetic near-field data

   https://doi.org/10.1515/jiip-2020-0114  

   Nguyen, Dinh-Liem; Truong, Trung (November 2020, Journal of Inverse and Ill-posed Problems)

   null (Ed.)

   Abstract This paper is concerned with the inverse scattering problem for the three-dimensional Maxwell equations in bi-anisotropic periodic structures.The inverse scattering problem aims to determine the shape of bi-anisotropic periodic scatterers from electromagnetic near-field data at a fixed frequency.The factorization method is studied as an analytical and numerical tool for solving the inverse problem.We provide a rigorous justification of the factorization method which results in the unique determination and a fast imaging algorithm for the periodic scatterer.Numerical examples for imaging three-dimensional periodic structures are presented to examine the efficiency of the method. 

   more » « less
2. End-to-End LU Factorization of Large Matrices on GPUs

   https://doi.org/10.1145/3572848.3577486  

   Xia, Yang; Jiang, Peng; Agrawal, Gagan; Ramnath, Rajiv (February 2023, PPoPP '23: Proceedings of the 28th ACM SIGPLAN Annual Symposium on Principles and Practice of Parallel Programming)

   LU factorization for sparse matrices is an important computing step for many engineering and scientific problems such as circuit simulation. There have been many efforts toward parallelizing and scaling this algorithm, which include the recent efforts targeting the GPUs. However, it is still challenging to deploy a complete sparse LU factorization workflow on a GPU due to high memory requirements and data dependencies. In this paper, we propose the first complete GPU solution for sparse LU factorization. To achieve this goal, we propose an out-of-core implementation of the symbolic execution phase, thus removing the bottleneck due to large intermediate data structures. Next, we propose a dynamic parallelism implementation of Kahn's algorithm for topological sort on the GPUs. Finally, for the numeric factorization phase, we increase the parallelism degree by removing the memory limits for large matrices as compared to the existing implementation approaches. Experimental results show that compared with an implementation modified from GLU 3.0, our out-of-core version achieves speedups of 1.13--32.65X. Further, our out-of-core implementation achieves a speedup of 1.2--2.2 over an optimized unified memory implementation on the GPU. Finally, we show that the optimizations we introduce for numeric factorization turn out to be effective. 

   more » « less
3. End-to-End LU Factorization of Large Matrices on GPUs

   Yang Xia, Peng Jiang (February 2023, Proceedings of PPoPP 2023)

   LU factorization for sparse matrices is an important computing step for many engineering and scientific problems such as circuit simulation. There have been many efforts toward parallelizing and scaling this algorithm, which include the recent efforts targeting the GPUs. However, it is still challenging to deploy a complete sparse LU factorization workflow on a GPU due to high memory requirements and data dependencies. In this paper, we propose the first complete GPU solution for sparse LU factorization. To achieve this goal, we propose an out-of-core implementation of the symbolic execution phase, thus removing the bottleneck due to large intermediate data structures. Next, we propose a dynamic parallelism implementation of Kahn's algorithm for topological sort on the GPUs. Finally, for the numeric factorization phase, we increase the parallelism degree by removing the memory limits for large matrices as compared to the existing implementation approaches. Experimental results show that compared with an implementation modified from GLU 3.0, our out-of-core version achieves speedups of 1.13--32.65X. Further, our out-of-core implementation achieves a speedup of 1.2--2.2 over an optimized unified memory implementation on the GPU. Finally, we show that the optimizations we introduce for numeric factorization turn out to be effective. 

   more » « less
4. Robust semi-supervised non-negative matrix factorization for binary subspace learning

   https://doi.org/10.1007/s40747-021-00285-1  

   Dai, Xiangguang; Zhang, Keke; Li, Juntang; Xiong, Jiang; Zhang, Nian; Li, Huaqing (February 2021, Complex & Intelligent Systems)

   null (Ed.)

   Abstract Non-negative matrix factorization and its extensions were applied to various areas (i.e., dimensionality reduction, clustering, etc.). When the original data are corrupted by outliers and noise, most of non-negative matrix factorization methods cannot achieve robust factorization and learn a subspace with binary codes. This paper puts forward a robust semi-supervised non-negative matrix factorization method for binary subspace learning, called RSNMF, for image clustering. For better clustering performance on the dataset contaminated by outliers and noise, we propose a weighted constraint on the noise matrix and impose manifold learning into non-negative matrix factorization. Moreover, we utilize the discrete hashing learning method to constrain the learned subspace, which can achieve a binary subspace from the original data. Experimental results validate the robustness and effectiveness of RSNMF in binary subspace learning and image clustering on the face dataset corrupted by Salt and Pepper noise and Contiguous Occlusion. 

   more » « less
5. Fresnel diffraction imaging of surface nanostructure using coherent resonant x-ray scattering

   https://doi.org/10.1063/5.0273192  

   Burgard, L.; Neupane, C.; Balodhi, A.; Bista, S.; Butun, S.; Jangid, R.; Barbour, A.; Basit, N.; Agterberg, D_F; Weinert, M.; et al (July 2025, Journal of Applied Physics)

   We investigated surface nanostructures on an antiferromagnet MnBi2Te4 using a novel imaging technique, direct (real)-space and real time coherent x-ray imaging (direct-CXI). This technique has provided new insights into antiferromagnetic textures, including the formation of anti-phase antiferromagnetic (AFM) domains and thermal dynamics of AFM domains and domain walls. While this method produces real-space images of AFM textures without requiring a complex imaging retrieval process, its underlying imaging mechanism has not been fully understood, limiting a deep understanding of AFM textures and the information they contain. By investigating the well-defined structural characteristics of the nanostructures fabricated on MnBi2Te4, we elucidate the imaging principle of this novel technique. We find that the observed images can be well explained by the Fresnel diffraction integral. Using a simple model from classical optics, our calculations successfully reproduce the experimentally observed images of the nanostructures. This demonstrates that direct-CXI not only provides straightforward real-space imaging but also contains phase information through its Fresnel diffraction integral. 

   more » « less