A Parallel Matrix Inversion Algorithm on Torus with Adaptive Pivoting
نویسندگان
چکیده
This paper presents a parallel algorithm for matrix inversion on a torus interconnected MIMD-MC multi-processor. This method is faster than the parallel implementations of other widely used methods namely Gauss-Jordan, Gauss-Seidal or LU decomposition based inversion. This new algorithm also introduces a novel technique, called adaptive pivoting, for solving the zero pivot problem at no cost. Our method eliminates the costly row interchange used by the existing elimination based parallel algorithms.This paper presents the design, analysisand simulation results (on a 32 Node Meiko Transputer) of this new and efficient matrix inversion algorithm.
منابع مشابه
A Parallel Method for Matrix Inversion Based on Gauss-jordan Algorithm
The inversion of matrix is widely used in many scientific applications and engineering calculations and this operation requires a large number of computational efforts and storage space. In order to reduce the consumed time and to increase efficiency, a parallel algorithm for matrix inversion based on classic Gauss-Jordan elimination with pivoting is proposed. Upon the multi-core DSPs platform,...
متن کاملAdaptive attitude controller of a reentry vehicles based on Back-stepping Dynamic inversion method
This paper presents an attitude control algorithm for a Reusable Launch Vehicle (RLV) with a low lift/drag ratio (L/D < 0.5), in the presence of external disturbances, model uncertainties, control output constraints and the thruster model. The main novelty of proposed control strategy is a new combination of the attitude control methods included backstepping, dynamic inversion and adaptive cont...
متن کاملScalable Stable Solvers for Non-symmetric Narrow-Banded Linear Systems
Banded linear systems with large bandwidths can be solved by similar methods as full linear systems. In particular, parallel algorithms based on torus-wrap mapping and Gaussian elimination with partial pivoting have been used with success. These algorithms are not suitable, however , if the bandwidth is small, say, between 1 and 100. As the bandwidth limits the amount of parallelism available a...
متن کاملAdaptive Algorithm-based Fault Tolerance for Parallel Computations in Linear Systems
This paper presents a dynamically adaptive stabilization scheme for parallel matrix computation. The scheme performs automatic error detection and correction through inserting redundant, but concurrent tracer computations within the folds of the regular computation. It also eliminates thecostly rowinterchangeused in classical pivoting. A fault-tolerant double wavefront matrix algorithm for a MI...
متن کاملMinimizing communication overhead for matrix inversion algorithms on hypercubes
The mirin contribution of this report is the development of novel algorithms {that make efficient use of the communication system in distributed memory architectures with plrocessing elements interconnected by a hypercube network. These algorithms achieve almost optirr~al overlap of communicatior~ delays by computation, leading to a minimization of communicatioi~ overhead. Rigorous ana1yt:ical ...
متن کامل