A sparse stiff chemistry solver based on dynamic adaptive integration for efficient combustion simulations

The sparse matrix solver has become the bottleneck in a Simulation Program with Integrated Circuit Emphasis circuit simulator. It is difficult to parallelize the sparse matrix solver because of the high data dependence during the numerical LU factorization. In this brief, a parallel LU factorization algorithm is developed on shared-memory computers with multicore central processing units, based on KLU algorithms. An Elimination Scheduler (EScheduler) is proposed to represent the data dependence during the LU factorization. Based on the EScheduler, the parallel tasks are scheduled in two modes to achieve a high level of concurrence, i.e., cluster mode and pipeline mode . The experimental results on 26 circuit matrices reveal that the developed algorithm can achieve speedup of 1.18-4.55t (on geometric average), as compared with KLU, with 1-8 threads. The result analysis indicates that for different data dependence, different parallel strategies should be dynamically selected to obtain optimal performance.

Radiation is an important process of energy transport, a force, and a basis for synthetic observations, so radiation hydrodynamics (RHD) calculations have occupied an important place in astrophysics. However, although the progress in computational technology is remarkable, their high numerical cost is still a persistent problem. In this work, we compare the following schemes used to solve the nonlinear simultaneous equations of an RHD algorithm with the flux-limited diffusion approximation: the Newton–Raphson (NR) method, operator splitting, and linearization (LIN), from the perspective of the computational cost involved. For operator splitting, in addition to the traditional simple operator splitting (SOS) scheme, we examined the scheme developed by Douglas & Rachford (DROS). We solve three test problems (the thermal relaxation mode, the relaxation and the propagation of linear waves, and radiating shock) using these schemes and then compare their dependence on the time step size. As a result, we find the conditions of the time step size necessary for adopting each scheme. The LIN scheme is superior to other schemes if the ratio of radiation pressure to gas pressure is sufficiently low. On the other hand, DROS can be the most efficient scheme if the ratio is high. Although the NR scheme can be adopted independently of the regime, especially in a problem that involves optically thin regions, the convergence tends to be worse. In all cases, SOS is not practical.

Assembling a coherent view of distributed heterogeneous information and their processing is challenging but important for inter-organizational business collaboration and service provision. However, traditional integration approaches do not consider dynamic and adaptive issues such as human intervention and exception handling. Therefore, we propose a Workflow-based Information Integration (WII) approach, which is particularly suitable in a loosely coupled Web services environment. Our implementation framework comprises five layers: semantic, application, workflow, service, and message. We focus on the workflow layer for providing adaptiveness from the aspects of various types of flows such as controlflows, data-flows, security-flows, exception-flows and semantic-flows by using the Business Process Execution Language for Web Services (BPEL). We further extend this with our proposed data-integration, semantic-referencing, and exception-handling assertions in order to achieve dynamic and adaptive workflow-based information integration plans. We map information into SOAP messages and link the proposed exception-handling assertions in BPEL to SOAP-fault implementations. We also define semantic referencing in BPEL by using OWL Web Ontology Language. Lastly, we demonstrate the feasibility of our adaptive approach with an intelligence information integration case study at the application layer and examine some typical use cases of exception-handling with semantic support.

Modern petroleum reservoir simulation serves as a primary tool for quantitatively managing reservoir production and planning new fields. It involves repeatedly solving the Jacobian of a set of strong nonlinear partial differential equations governing the mass and energy conduction and conservation. Most of the existing reservoir simulators adopt iterative solver with multiple stages of preconditioners, in which the incomplete LU (ILU) factorization is an outstanding universal smoother. However, it turns out that when the degree of freedom of each grid grows, ILU usually becomes the bottleneck of the solver. Moreover, ILU is difficult to parallelize due to its inherent data dependency. In this paper, we developed a sparse iterative solver with parallelized ILU and triangular solve using block-wise data structure. Compared with the state of art iterative solver on 14 industrial reservoir simulation matrices, the proposed ILU is 5.2x faster (on average) than the state of art iterative solver because of the block-wise data structure, which leads to 2.2x speedup on the total solver runtime. In addition, parallel ILU and triangular solve are developed to further accelerate the solver. To tackle the strong data dependency in ILU and triangular solve, we first partition the algorithm into separated tasks and construct a data flow graph to represent the data dependency. Then, tasks are scheduled in parallel according to the topological order of the data flow graph. On an 8-thread multicore architecture, we achieved another 3.6x speedup on ILU factorization, and 3.3x on triangular solve with good scalability.

In this paper, we present a novel approach to human-robot control. Taking inspiration from behaviour-based robotics and self-organisation principles, we present an interfacing mechanism, with the ability to adapt both towards the user and the robotic morphology. The aim is for a transparent mechanism connecting user and robot, allowing for a seamless integration of control signals and robot behaviours. Instead of the user adapting to the interface and control paradigm, the proposed architecture allows the user to shape the control motifs in their way of preference, moving away from the case where the user has to read and understand an operation manual, or it has to learn to operate a specific device. Starting from a tabula rasa basis, the architecture is able to identify control patterns (behaviours) for the given robotic morphology and successfully merge them with control signals from the user, regardless of the input device used. The structural components of the interface are presented and assessed both individually and as a whole. Inherent properties of the architecture are presented and explained. At the same time, emergent properties are presented and investigated. As a whole, this paradigm of control is found to highlight the potential for a change in the paradigm of robotic control, and a new level in the taxonomy of human in the loop systems.

A GENERALIZED REDUCED GRADIENT ALGORITHM FOR SOLVING LARGE-SCALE DISCRETE-TIME NONLINEAR OPTIMAL CONTROL PROBLEMS

The effectiveness of sparse matrix techniques for directly solving large-scale linear least-squares problems is severely limited if the system matrix A has one or more nearly dense rows. In this paper, we partition the rows of A into sparse rows and dense rows (As and Ad) and apply the Schur complement approach. A potential difficulty is that the reduced normal matrix AsTAs is often rank-deficient, even if A is of full rank. To overcome this, we propose explicitly removing null columns of As and then employing a regularization parameter and using the resulting Cholesky factors as a preconditioner for an iterative solver applied to the symmetric indefinite reduced augmented system. We consider complete factorizations as well as incomplete Cholesky factorizations of the shifted reduced normal matrix. Numerical experiments are performed on a range of large least-squares problems arising from practical applications. These demonstrate the effectiveness of the proposed approach when combined with either a sparse parallel direct solver or a robust incomplete Cholesky factorization algorithm.

This letter presents techniques for tuning an accelerated preconditioned conjugate gradient solver with a multilevel preconditioner. The solver is optimized for a fast solution of sparse systems of equations arising in computational electromagnetics in a finite element method using higher-order elements. The goal of the tuning is to increase the throughput while at the same time reducing the memory requirements in order to allow one to process very large complex or real systems in single and double precision using commodity graphic processing units (GPUs). A threefold memory footprint reduction is achieved by means of a new format of storing sparse matrices. The acceleration is achieved by optimizing a sparse matrix-vector product on a GPU by applying new features of the Fermi architecture. Further improvements are obtained by introducing more levels into the preconditioner and the application of a fast sparse direct solver for the operations executed on a CPU. Numerical results for a setup consisting of a Fermi GPU (GTX 480) and a Xeon six-core CPU showed that the proposed approach allows one to handle systems involving millions of unknowns and reach the speedup factor of almost 4 compared to the CPU-only implementation.

This work is focused on analyzing potential performance improvements of HPC applications using stacked memories like the Hybrid Memory Cube, or HMC. We target a HPC sparse direct solver library, SuperLU [4], that performs LU decomposition and is a core piece of simulation codes like NIMROD [1]. To accelerate this library, we are interested in mapping both the computationally intense Spare Matrix-Vector (SpMV) kernels that can be implemented using matrix-matrix multiply (GEMM) calls and memory-intensive primitives like Scatter and Gather to a reconfigurable fabric tightly integrated with a 3D stacked memory. Here we provide initial results on mapping GEMM to OpenCL-based devices as well as a trace-driven evaluation of SuperLU's memory accesses with a combined FPGA and HMC platform.

We propose a scalable, and memory and time efficient parallel blockwise-sparse LU decomposition for distributed NUMA architectures suitable for global domain decomposition preconditioning. A static task scheduling heuristic is used to spatial and temporary allocate computational tasks on distributed NUMA systems. Asynchronous communication aided by manual progression techniques are used to maximize communication and computation overlap. Comparisons with state of the art sparse direct solvers on various electrically large problems suggest up to 2 times less memory and up to 4 times faster time while having better parallel efficiency.

The main purpose of this work is to present a new parallel direct solver: Dissection solver. It is based on LU factorization of the sparse matrix of the linear system and allows to detect automatically and handle properly the zero‐energy modes, which are important when dealing with DDM. A performance evaluation and comparisons with other direct solvers (MUMPS, DSCPACK) are also given for both sequential and parallel computations. Results of numerical experiments with a two‐level parallelization of large‐scale structural analysis problems are also presented: FETI is used for the global problem parallelization and Dissection for the local multithreading. In this framework, the largest problem we have solved is of an elastic solid composed of 400 subdomains running on 400 computation nodes (3200 cores) and containing about 165 millions dof. The computation of one single iteration consumes less than 20 min of CPU time. Several comparisons to MUMPS are given for the numerical computation of large‐scale linear systems on a massively parallel cluster: performances and weaknesses of this new solver are highlighted. Copyright © 2011 John Wiley & Sons, Ltd.

Discontinuous Galerkin method for multicomponent chemically reacting flows and combustion

The explosive growth of social media platforms has generated an unprecedented volume of user-generated content, making sentiment classification a crucial task for understanding public opinion, brand perception, and societal trends. However, the informal language, context-dependent expressions, and sequential dependencies in social media texts pose significant challenges for conventional machine learning and deep learning models. To address these challenges, this study proposes a Hybrid Attention-Driven Recurrent Neural Network (HADRNN) model for sentiment classification of social media texts. The model integrates bidirectional gated recurrent units (Bi-GRUs) with a multi-level attention mechanism, enabling it to effectively capture both global semantic context and fine-grained sentiment cues. Pre-trained GloVe embeddings are employed for robust lexical representation, while convolutional layers extract local n-gram features that complement the sequential modeling capabilities of recurrent units. The attention module adaptively emphasizes sentiment-bearing words and contextual phrases, ensuring both interpretability and improved performance. Extensive experiments were conducted on benchmark datasets including Twitter Sentiment140 and IMDB reviews. The proposed HADRNN achieved an accuracy of 92.1%, F1-score of 91.3%, precision of 91.7%, and recall of 90.8% on Twitter Sentiment140, significantly outperforming baseline CNN (87.2% accuracy), LSTM (88.9% accuracy), and Transformer-based models (90.4% accuracy). On IMDB, the model obtained an accuracy of 94.6% and F1-score of 94.1%, demonstrating robustness to longer text sequences and noisy language. These findings confirm that the hybrid integration of recurrent dynamics, attention, and convolutional processing not only enhances classification performance but also provides interpretable insights into sentiment-laden expressions, offering a reliable framework for real-world social media analytics.

In mixed quantum-classical molecular dynamics few but important degrees of freedom of a dynamical system are modeled quantum-mechanically while the remaining ones are treated within the classical approximation. Rothe methods established in the theory of partial differential equations are used to control both temporal and spatial discretization errors on grounds of a global tolerance criterion. The TRAIL (trapezoidal rule for adaptive integration of Liouville dynamics) scheme [I. Horenko and M. Weiser, J. Comput. Chem. 24, 1921 (2003)] has been extended to account for nonadiabatic effects in molecular dynamics described by the quantum-classical Liouville equation. In the context of particle methods, the quality of the spatial approximation of the phase-space distributions is maximized while the numerical condition of the least-squares problem for the parameters of particles is minimized. The resulting dynamical scheme is based on a simultaneous propagation of moving particles (Gaussian and Dirac deltalike trajectories) in phase space employing a fully adaptive strategy to upgrade Dirac to Gaussian particles and, vice versa, downgrading Gaussians to Dirac-type trajectories. This allows for the combination of Monte-Carlo-based strategies for the sampling of densities and coherences in multidimensional problems with deterministic treatment of nonadiabatic effects. Numerical examples demonstrate the application of the method to spin-boson systems in different dimensionality. Nonadiabatic effects occurring at conical intersections are treated in the diabatic representation. By decreasing the global tolerance, the numerical solution obtained from the TRAIL scheme are shown to converge towards exact results.

An evaluation of temporally adaptive transformation approaches for solving Richards' equation