Topic
Synchronization (computer science)
About: Synchronization (computer science) is a research topic. Over the lifetime, 43303 publications have been published within this topic receiving 559005 citations.
Papers published on a yearly basis
Papers
More filters
01 Jan 1989
TL;DR: This book focuses on numerical algorithms suited for parallelization for solving systems of equations and optimization problems, with emphasis on relaxation methods of the Jacobi and Gauss-Seidel type.
Abstract: This book focuses on numerical algorithms suited for parallelization for solving systems of equations and optimization problems Emphasis on relaxation methods of the Jacobi and Gauss-Seidel type, and issues of communication and synchronization Topics covered include: Algorithms for systems of linear equations and matrix inversion; Herative methods for nonlinear problems; and Shortest paths and dynamic programming
1,423 citations
Posted Content•
TL;DR: This work aims to show using novel theoretical analysis, algorithms, and implementation that SGD can be implemented without any locking, and presents an update scheme called HOGWILD! which allows processors access to shared memory with the possibility of overwriting each other's work.
Abstract: Stochastic Gradient Descent (SGD) is a popular algorithm that can achieve state-of-the-art performance on a variety of machine learning tasks. Several researchers have recently proposed schemes to parallelize SGD, but all require performance-destroying memory locking and synchronization. This work aims to show using novel theoretical analysis, algorithms, and implementation that SGD can be implemented without any locking. We present an update scheme called HOGWILD! which allows processors access to shared memory with the possibility of overwriting each other's work. We show that when the associated optimization problem is sparse, meaning most gradient updates only modify small parts of the decision variable, then HOGWILD! achieves a nearly optimal rate of convergence. We demonstrate experimentally that HOGWILD! outperforms alternative schemes that use locking by an order of magnitude.
1,413 citations
01 Jan 2008
TL;DR: In this article, the synchronization skeleton of a program is synthesized from a high-level Temporal Logic specification, which is an abstraction of the actual program where detail irrelevant to synchronization is suppressed.
Abstract: We Propose a method of constructing concurrent programs in which the synchronization skeletonof the program is automatically synthesized from a high-level (branching time) Temporal Logic specification. The synchronization skeleton is an abstraction of the actual program where detail irrelevant to synchronization is suppressed. For example, in the synchronization skeleton for a solution to the critical section problem each process's critical section may be viewed as a single node since the internal structure of the critical section is unimportant. Most solutions to synchronization problems in the literature are in fact given as synchronization skeletons. Because synchronization skeletons are in general finite state, the propositional version of Temporal Logic can be used to specify their properties.
1,399 citations
TL;DR: This work proposes that processes underlying working and long-term memory might interact in the medial temporal lobe and proposes that this is accomplished by neural operations involving phase–phase and phase–amplitude synchronization.
Abstract: In recent years, studies ranging from single-unit recordings in animals to electroencephalography and magnetoencephalography studies in humans have demonstrated the pivotal role of phase synchronization in memory processes. Phase synchronization - here referring to the synchronization of oscillatory phases between different brain regions - supports both working memory and long-term memory and acts by facilitating neural communication and by promoting neural plasticity. There is evidence that processes underlying working and long-term memory might interact in the medial temporal lobe. We propose that this is accomplished by neural operations involving phase-phase and phase-amplitude synchronization. A deeper understanding of how phase synchronization supports the flexibility of and interaction between memory systems may yield new insights into the functions of phase synchronization in general.
1,396 citations
TL;DR: In this paper, the authors apply the concept of phase synchronization of chaotic and/or noisy systems and the statistical distribution of the relative instantaneous phases to electroencephalograms (EEGs) recorded from patients with temporal lobe epilepsy.
1,232 citations