Accelerating Numerical Linear Algebra Kernels on a Scalable Run Time Reconfigurable Platform

Numerical Linear Algebra (NLA) kernels are at the heart of all computational problems. These kernels require hardware acceleration for increased throughput. NLA Solvers for dense and sparse matrices differ in the way the matrices are stored and operated upon although they exhibit similar computational properties. While ASIC solutions for NLA Solvers can deliver high performance, they are not scalable, and hence are not commercially viable. In this paper, we show how NLA kernels can be accelerated on REDEFINE, a scalable runtime reconfigurable hardware platform. Compared to a software implementation, Direct Solver (Modified Faddeev's algorithm) on REDEFINE shows a 29X improvement on an average and Iterative Solver (Conjugate Gradient algorithm) shows a 15-20% improvement. We further show that solution on REDEFINE is scalable over larger problem sizes without any notable degradation in performance.

Topological analysis of general linear networks

A new method of network analysis, a generalization in several different senses of existing methods and applicable to all networks for which a branch-admittance (or impedance) matrix can be formed, is presented. The treatment of network determinants is very general and essentially four terminal rather than three terminal, and leads to simple expressions based on trees of a simple graph associated with the network and matrix, and involving products of low-order, usually(2 times 2)determinants of tree-branch admittances, in addition to tree-branch products as in existing methods. By comparison with existing methods, the total number of trees and of tree pairs is usually considerably reduced, and this fact, together with an easy method of tree-pair sign determination which is also presented, makes the new method simpler in general. The method can be very easily adapted, by the use of infinite parameters, to accommodate ideal transformers, operational amplifiers, and other forms of network constraint; in fact, is thought to be applicable to all linear networks.

Class of linear time-varying discrete systems

A class of linear time-varying discrete systems is considered, and closed-form solutions are obtained in different cases. Some comments on stability are also included.

Study of symmetrical and related components through the theory of linear vector spaces

The paper proposes a study of symmetrical and related components, based on the theory of linear vector spaces. Using the concept of equivalence, the transformation matrixes of Clarke, Kimbark, Concordia, Boyajian and Koga are shown to be column equivalent to Fortescue's symmetrical-component transformation matrix. With a constraint on power, criteria are presented for the choice of bases for voltage and current vector spaces. In particular, it is shown that, for power invariance, either the same orthonormal (self-reciprocal) basis must be chosen for both voltage and current vector spaces, or the basis of one must be chosen to be reciprocal to that of the other. The original �¿, ��, 0 components of Clarke are modified to achieve power invariance. For machine analysis, it is shown that invariant transformations lead to reciprocal mutual inductances between the equivalent circuits. The relative merits of the various components are discussed.

Matrix Processors Using p-adic Arithmetic for Exact Linear Computations

A unique code (called Hensel's code) is derived for a rational number by truncating its infinite p-adic expansion. The four basic arithmetic algorithms for these codes are described and their application to rational matrix computations is demonstrated by solving a system of linear equations exactly, using the Gaussian elimination procedure.

Control technique for non-linear and unbalance compensation of an integrated magnetics-based three-phase series-shunt-compensated uninterruptible power supply

Handling unbalanced and non-linear loads in a three-phase AC power supply has always been a difficult issue. This has been addressed in the literature by either using fast controllers in the fundamental rotating reference frame or using separate controllers in reference frames specific to the harmonics. In the former case, the controller needs to be fast and in the latter case, besides the need for many controllers, negative-sequence components need to be extracted from the measured signal. This study proposes a control scheme for harmonic and unbalance compensation of a three-phase uninterruptible power supply wherein the problems mentioned above are addressed. The control takes place in the fundamental positive-sequence reference frame using only a set of feedback and feed-forward compensators. The harmonic components are extracted by a process of frame transformations and used as feed-forward compensation terms in the positive-sequence fundamental reference frame. This study uses a method wherein the measured signal itself is used for fundamental negative-sequence compensation. As the feed-forward compensator handles the high-bandwidth components, the feedback compensator can be a simple low-bandwidth one. This control algorithm is explained and validated experimentally.

Design of linear systems for decoupling and disturbance rejection

The question of achieving decoupling and asymptotic disturbance rejection in time-invariant linear multivariable systems subject to unmeasurable arbitrary disturbances of a given class is discussed. A synthesis procedure which determines a feedback structure, incorporating an integral compensator, is presented.

Distributed Function Computation over Fields and Rings via Linear Compression of Sources

The setting considered in this paper is one of distributed function computation. More specifically, there is a collection of N sources possessing correlated information and a destination that would like to acquire a specific linear combination of the N sources. We address both the case when the common alphabet of the sources is a finite field and the case when it is a finite, commutative principal ideal ring with identity. The goal is to minimize the total amount of information needed to be transmitted by the N sources while enabling reliable recovery at the destination of the linear combination sought. One means of achieving this goal is for each of the sources to compress all the information it possesses and transmit this to the receiver. The Slepian-Wolf theorem of information theory governs the minimum rate at which each source must transmit while enabling all data to be reliably recovered at the receiver. However, recovering all the data at the destination is often wasteful of resources since the destination is only interested in computing a specific linear combination. An alternative explored here is one in which each source is compressed using a common linear mapping and then transmitted to the destination which then proceeds to use linearity to directly recover the needed linear combination. The article is part review and presents in part, new results. The portion of the paper that deals with finite fields is previously known material, while that dealing with rings is mostly new.Attempting to find the best linear map that will enable function computation forces us to consider the linear compression of source. While in the finite field case, it is known that a source can be linearly compressed down to its entropy, it turns out that the same does not hold in the case of rings. An explanation for this curious interplay between algebra and information theory is also provided in this paper.

Control technique for non-linear and unbalance compensation of an integrated magnetics-based three-phase series-shunt-compensated uninterruptible power supply

Handling unbalanced and non-linear loads in a three-phase AC power supply has always been a difficult issue. This has been addressed in the literature by either using fast controllers in the fundamental rotating reference frame or using separate controllers in reference frames specific to the harmonics. In the former case, the controller needs to be fast and in the latter case, besides the need for many controllers, negative-sequence components need to be extracted from the measured signal. This study proposes a control scheme for harmonic and unbalance compensation of a three-phase uninterruptible power supply wherein the problems mentioned above are addressed. The control takes place in the fundamental positive-sequence reference frame using only a set of feedback and feed-forward compensators. The harmonic components are extracted by a process of frame transformations and used as feed-forward compensation terms in the positive-sequence fundamental reference frame. This study uses a method wherein the measured signal itself is used for fundamental negative-sequence compensation. As the feed-forward compensator handles the high-bandwidth components, the feedback compensator can be a simple low-bandwidth one. This control algorithm is explained and validated experimentally.

Control technique for non-linear and unbalance compensation of an integrated magnetics-based three-phase series-shunt-compensated uninterruptible power supply

Handling unbalanced and non-linear loads in a three-phase AC power supply has always been a difficult issue. This has been addressed in the literature by either using fast controllers in the fundamental rotating reference frame or using separate controllers in reference frames specific to the harmonics. In the former case, the controller needs to be fast and in the lattercase, besides the need for many controllers, negative-sequence components need to be extracted from the measured signal.This study proposes a control scheme for harmonic and unbalance compensation of a three-phase uninterruptible power supply wherein the problems mentioned above are addressed. The control takes place in the fundamental positive-sequence reference frame using only a set of feedback and feed-forward compensators. The harmonic components are extracted by process of frame transformations and used as feed-forward compensation terms in the positive-sequence fundamental reference frame. This study uses a method wherein the measured signal itself is used for fundamental negative-sequence compensation. As the feed-forward compensator handles the high-bandwidth components, the feedback compensator can be a simple low-bandwidth one. This control algorithm is explained and validated experimentally.

2.5PN linear momentum flux from inspiralling compact binaries in quasicircular orbits and associated recoil: Nonspinning case

Anisotropic emission of gravitational waves (GWs) from inspiralling compact binaries leads to the loss of linear momentum and hence gravitational recoil of the system. The loss rate of linear momentum in the far-zone of the source (a nonspinning binary system of black holes in quasicircular orbit) is investigated at the 2.5 post-Newtonian (PN) order and used to provide an analytical expression in harmonic coordinates for the 2.5PN accurate recoil velocity of the binary accumulated in the inspiral phase. The maximum recoil velocity of the binary system at the end of its inspiral phase (i.e at the innermost stable circular orbit (ISCO)) estimated by the 2.5PN formula is of the order of 4 km s(-1) which is smaller than the 2PN estimate of 22 km s(-1). Going beyond inspiral, we also provide an estimate of the more important contribution to the recoil velocity from the plunge phase. The maximum recoil velocity at the end of the plunge, involving contributions both from inspiral and plunge phase, for a binary with symmetric mass ratio nu = 0.2 is of the order of 182 km s(-1).

Non-linear electrical response of chalcogenide glasses: Memory state phenomena

Investigations on the switching behaviour of arsenic-tellurium glasses with Ge or Al additives, yield interesting information about the dependence of switching on network rigidity, co-ordination of the constituents, glass transition & ambient temperature and glass forming ability.

Structural Transformations, Composition Anomalies and a Dramatic Collapse of Linear Polymer Chains in Dilute Ethanol-Water Mixtures

Water-ethanol mixtures exhibit many interesting anomalies, such as negative excess partial molar volume of ethanol, excess sound absorption coefficient at low concentrations, and positive deviation from Raoult's law for vapor pressure, to mention a few. These anomalies have been attributed to different, often contradictory origins, but a quantitative understanding is still lacking. We show by computer simulation and theoretical analyses that these anomalies arise from the sudden emergence of a bicontinuous phase that occurs at a relatively low ethanol concentration of x(eth) approximate to 0.06-0.10 (that amounts to a volume fraction of 0.17-0.26, which is a significant range!). The bicontinuous phase is formed by aggregation of ethanol molecules, resulting in a weak phase transition whose nature is elucidated. We find that the microheterogeneous structure of the mixture gives rise to a pronounced nonmonotonic composition dependence of local compressibility and nonmonotonic dependence in the peak value of the radial distribution function of ethyl groups. A multidimensional free energy surface of pair association is shown to provide a molecular explanation of the known negative excess partial volume of ethanol in terms of parallel orientation and hence better packing of the ethyl groups in the mixture due to hydrophobic interactions. The energy distribution of the ethanol molecules indicates additional energy decay channels that explain the excess sound attenuation coefficient in aqueous alcohol mixtures. We studied the dependence of the solvation of a linear polymer chain on the composition of the water-ethanol solvent. We find that there is a sudden collapse of the polymer at x(eth) approximate to 0.05-a phenomenon which we attribute to the formation of the microheterogeneous structures in the binary mixture at low ethanol concentrations. Together with recent single molecule pulling experiments, these results provide new insight into the behavior of polymer chain and foreign solutes, such as enzymes, in aqueous binary mixtures.

The canonic linear-phase FIR lattice structures

The Linear phase(LP) Finite Impulse Response(FIR) filters are widely used in many signal processing systems which are sensitive to phase distortion. In this article, we obtain a canonic lattice structure of an LP-FIR filter with a complex impulse response. This lattice structure is based on some novel lattice stages obtained from some properties of symmetric polynomials.This canonic lattice structure exploits the redundancy in the zeros of an LP-FIR filter.