Achieving full parallelism using multidimensional retiming

Most scientific and digital signal processing (DSP) applications are recursive or iterative. Transformation techniques are usually applied to get optimal execution rates in parallel and/or pipeline systems. The retiming technique is a common and valuable transformation tool in one-dimensional problems, when loops are represented by data flow graphs (DFGs). In this paper, uniform nested loops are modeled as multidimensional data flow graphs (MDFGs). Full parallelism of the loop body, i.e., all nodes in the MDFG executed in parallel, substantially decreases the overall computation time. It is well known that, for one-dimensional DFGs, retiming can not always achieve full parallelism. Other existing optimization techniques for nested loops also can not always achieve full parallelism. This paper shows an important and counter-intuitive result, which proves that we can always obtain full-parallelism for MDFGs with more than one dimension. This result is obtained by transforming the MDFG into a new structure. The restructuring process is based on a multidimensional retiming technique. The theory and two algorithms to obtain full parallelism are presented in this paper. Examples of optimization of nested loops and digital signal processing designs are shown to demonstrate the effectiveness of the algorithms     

Construction of maximum cycles in faulty binary hypercubes

Consideration was given to the following problem. In the binary hypercube, given is a Hamiltonian cycle with faulty edges, or vertices, or both. Needed is to construct a length-maximum cycle without faulty components of the hypercube. The cycles are defined by the ring sequences of the weights of the hypercube edges belonging to them. The discussion was based on the example of a binary 4-dimensional hypercube.Translated from Avtomatika i Telemekhanika, No. 4, 2005, pp. 141–155.Original Russian Text Copyright © 2005 by Parkhomenko.This paper was recommended for publication by P.Yu. Chebotarev, a member of the Editorial Board     

Variables Bounding Based Retiming Algorithm

Retiming is a technique for optimizing sequential circuits.In this paper,we discuss this problem and propose an improved retiming algorithm based on varialbes bounding.Through the computation of the lower and upper bounds on variables,the algorithm can significantly reduce the number of constratints and speed up the execution of retiming.Furthermore,the elements of matrixes D and W are computed in a demand-driven way,which can reduce the capacity of memory,It is shown through the experimental results on ISCAS89 benchmarks that our algorithm is very effective for large-scale seuqential circuits.     

Using fixed-point semantics to prove retiming lemmas

Algorithms designed for VLSI implementation are commonly described by directed graphs, in which the nodes represent functional units and the arcs indicate communication links. We give a denotational semantics for such a graph in terms of the least fixed point of a set of (mutually recursive) function definitions, describing the outputs produced at each node as a function of time. This semantics is consistent with the conventional clocked operational semantics of the system. A retiming is a systematic modification of the internode delays of a design, often used to convert an algorithm design into a systolic form. The utility of such retimings in optimizing the behavior of designs is well known. We use fixed-point semantics to provide simple proofs of the correctness of certain retiming transformations from the literature and to justify other design transformations such as pipelining.     

Understanding language through vision

Artificial Intelligence Review - We discuss briefly (from a philosophical point of view) why interfacing a linguistic analyzer with an artificial vision system is an important issue. In particular...     

Scheduling data-flow graphs via retiming and unfolding

Loop scheduling is an important problem in parallel processing. The retiming technique reorganizes an iteration; the unfolding technique schedules several iterations together. We combine these two techniques to obtain a static schedule with a reduced average computation time per iteration. We first prove that the order of retiming and unfolding is immaterial for scheduling a data-flow graph (DFG). From this nice property, we present a polynomial-time algorithm on the original DFG, before unfolding, to find the minimum-rate static schedule for a given unfolding factor. For the case of a unit-time DFG, efficient checking and retiming algorithms are presented     

Circuit retiming applied to decomposed software pipelining

This paper elaborates on a new view on software pipelining, called decomposed software pipelining. The approach is to decouple the problem into resource constraints and dependence constraints. Resource constraints management amounts to scheduling an acyclic graph subject to resource constraints for which an efficiency bound is known, resulting in a bound for loop scheduling. The acyclic graph is obtained by cutting some particular edges of the (cyclic) dependence graph. In this paper, we cut edges in a different way, using circuit retiming algorithms, so as to minimize both the longest dependence path in the acyclic graph, and the number of edges in the acyclic graph. With this technique, we improve the efficiency bound given for Gasperoni and Schwlegelshohn algorithm, and we reduce the constraints that remain for the acyclic problem. We believe this framework to be of interest because it brings a new insight into the software problem by establishing its deep link with the circuit retiming problem     

Better Understanding through Formal Specification

The Graphical Kernel System (GKS) is now registered as an ISO International Standard for computer graphics programming. One of the major innovations of the Standard is the bundled specification of aspects, a mechanism which gives the applications programmer the ability to tailor the appearance of a picture independently on each of the workstations on which it is displayed, using the capabilities of the workstations. GKS also incorporates the traditional method of individual specification of aspects in which each workstation does the best it can to represent global aspect values. In this paper a formal specification technique, the Vienna Development Method (VDM), is used to describe aspect specification. The GKS model of aspect specification is progressively constructed from simpler models. Properties of these simpler models are formulated and the specifications are proved to conform to these. The properties are then traced through the more complex models. The paper demonstrates the applicability of formal specification to the design of graphics software and the ability of formal techniques to catalyse the deeper understanding of designs.     

Understanding Cognition through Synthesis and Analysis

This article discusses the interdisciplinary research area of cognitive science. It presents methods and approaches to investigating cognitive processes. Spatial cognition as a subfield of cognitive science is used to show how mental processes are connected with the physical realm through perception and action. These bridges between physical and mental permit a variety of rather distinct approaches that contribute to the investigation of the cognitive processes involved. Particular attention is given to the role of artificial intelligence approaches to help discovering cognitive processing principles. The article discusses the interplay between synthetic and analytic approaches to cognitive science and emphasizes the complementary strengths of these approaches. It shows how the environment, the context of the specific situation, and the task can reduce the cognitive effort to be taken. The trade-off between general abstract representations and specific concrete representations is shown. The power of virtual environments for testing and understanding models of cognition is demonstrated. Numerous examples are given.     

Understanding design activities through computer simulation

In this paper, a new experimental approach – computer simulation – is introduced for understanding design activities and for validating design theories. Following the generic framework of computer simulation, three main components in simulating design activities are introduced: mathematical model, simulation model, and statistical analysis. The mathematical model consists of the design governing equation and Environment-Based Design (EBD), based on which three routes are introduced to look for new design solutions: (1) formulating the design problem differently at the beginning of a design process may get quite different solutions, in which creative design could emerge; (2) extending designer’s knowledge and experience can help generate more candidate solutions, and so increasing the probability of generating a good concept; (3) changing the sequence of design problem decomposition may change product requirements, and thus change the generated design concepts. By viewing mesh generation algorithms as design agents, a computer simulation environment is used to study design activities. Statistical analysis is conducted to validate quantitatively the three routes to new design solutions. The results show that computer simulation is an effective approach to studying design activities.     

GRAFLOG: Understanding Drawings through Natural Language

This paper describes an experimental interactive graphics interface, GRAFLOG, in which drawings receive linguistic interpretations. It is possible to emulate linguistic interaction in situations where graphics is thought to be necessary. The paper presents examples of such a kind of dialogue and the architecture of the implementation. The paper explains how representations of drawings can be constructed by treating graphical symbols as "objects", and how a parallel linguistic interpretation for these symbols can be constructed. It highlights the relevance of "deictic expressions" and "spatial prepositions" in building the interface mechanisms between these two kinds of representations. Lastly, it shows how a reasoning component is constructed for making deductions from premises that are found in both the graphical and linguistic domains. Using GRAFLOG, it is possible to represent knowledge through words and pictures. GRAFLOG is implemented, using an object oriented programming style, in PROLOG and GKS.     

An integrated aircraft routing,crew scheduling and flight retiming model

In the integrated aircraft routing, crew scheduling and flight retiming problem, a minimum-cost set of aircraft routes and crew pairings must be constructed while choosing a departure time for each flight leg within a given time window. Linking constraints ensure that the same schedule is chosen for both the aircraft routes and the crew pairings, and impose minimum connection times for crews that depend on aircraft connections and departure times. We propose a compact formulation of the problem and a Benders decomposition method with a dynamic constraint generation procedure to solve it. Computational experiments performed on test instances provided by two major airlines show that allowing some flexibility on the departure times within an integrated model yields significant cost savings while ensuring the feasibility of the resulting aircraft routes and crew pairings.     

Increasing Understanding of Technology Management through Challenge Mapping

The authors seek to clarify what is understood by the management of technology by involving a group of managers in a workshop to identify specifically what they believe to be their critical difficulties and challenges in managing technology. The approach is shown to be a means of enhancing creativity, designed not so much for creating good answers and solutions, but rather for discovering good questions and challenges and for conceptualizing complex issues. Virtually all of the challenges selected as most important had much more to do with leading people to think more innovatively and collaboratively than with new technology itself.     

Understanding preservice teachers' technology use through TPACK framework

This study discusses preservice teachers' achievement barriers to technology integration, using principles of technological pedagogical content knowledge (TPACK) as an evaluative framework. Technology‐capable participants each freely chose a content area to comprise project. Data analysis based on interactions among core components of TPACK revealed that participants struggled with developing new knowledge. Lack of pedagogical experience limited development of appropriate technology integration approaches. Creating new knowledge bases based on different teaching components can be difficult for preservice teachers because it requires a deep understanding of core knowledge and interpretation of the teaching context and its dynamics. Developing pedagogical content knowledge (PCK) is an important factor in overall technology integration; teachers must make it a priority to acquire PCK before integrating technology. In preservice teacher education, PCK development must be supported with actual teaching experience. We believe that the results of the study may provide valuable insight with respect to proper focus on technology integration and recognizing limitations and challenges within TPACK principles to both those who teach technology integration and those who design TPACK‐based activities.     

Understanding topological relationships through comparisons of similar knots

This paper examines an example of learning with artifacts using the commonplace materials of string and knots. Emphases include research into learning processes as well as construction of objects to assist learning. The inquiry concerns the development of mathematical thinking, topology in particular. The research methodology combines participant observation and clinical interview within a constructionist framework. The study was set in a self-styled, self-constructed environment that consisted of knots and a social substrate encouraging lively exchanges of ideas about them. Comparisons of certain knots helped to elicit conceptions of the fundamental topological relationships of neighborhood, continuity, and boundaries. The paper includes comments on the suitability of specific artifacts for specific kinds of thinking and learning, and emphasizes the importance for software design of considering different learning styles.     

Theoretical comparison between sequential redundancy addition and removal and retiming optimization techniques

This paper attempts to determine the capabilities of existing redundancy addition and removal (SRAR) techniques for logic optimization of sequential circuits. To this purpose, we compare this method with the retiming and resynthesis (RaR) techniques. For the RaR case the set of possible transformations has been established by relating them to STG transformations by other authors. Following these works, we first formally demonstrate that logic transformations provided by RaR are covered by SRAR as well. Then we also show that SRAR is able to identify transformations that cannot be found by RaR. This way we prove that the sequential redundancy addition and removal technique provides more possibilities for logic optimization.     

Hamiltonian paths and cycles passing through a prescribed path in hypercubes

Assume that P is any path in a bipartite graph G of length k with 2?k?h, G is said to be h-path bipancyclic if there exists a cycle C in G of every even length from 2k to |V(G)| such that P lies in C. In this paper, the following result is obtained: The n-dimensional hypercube Qn with n?3 is (2n−3)-path bipancyclic but is not (2n−2)-path bipancyclic, moreover, a path P of length k with 2?k?2n−3 lies in a cycle of length 2k−2 if and only if P contains two edges of the same dimension. In order to prove the above result we first show that any path of length at most 2n−1 is a subpath of a Hamiltonian path in Qn with n?2, moreover, the upper bound 2n−1 is sharp when n?4.     

Understanding TRIZ through the review of top cited publications

The development of the Theory of Inventive Problem Solving (TRIZ) has not followed the usual patterns of scientific validation required by engineering methods. Consequently, its outreach within engineering design is interpreted differently in the scholarly community. At the same time, the claimed powerful support in tackling technical problems of any degree of difficulty conflicts with TRIZ diffusion in industrial settings, which is relatively low according to insights into product development practices. The mismatch between ambitious goals and moderate spill-over benefits in the industry ranges among the various open issues concerning TRIZ, its way of thinking, its effectiveness, the usability of its tools. In order to provide a general overview of TRIZ in science, the authors have attempted to analyse reliable and influential sources from the literature. The performed survey includes the top 100 indexed publications concerning TRIZ, according to the number of received citations. Variegated and poorly interconnected research directions emerge in the abundant literature that tackles TRIZ-related topics. The outcomes of the investigation highlight the successful implementation of TRIZ within, among the others, biomimetics and information processing. The traditional borders of mechanical and industrial engineering have been frequently crossed, as the use of TRIZ is also witnessed in the domain of business and services. At the same time, computer-aided platforms represent diffused attempts to boost TRIZ diffusion and applicability.     

Understanding user intent on the web through interaction mining

Predicting the goals of internet users can be extremely useful in e-commerce, online entertainment, and many other internet-based applications. One of the crucial steps to achieve this is to classify internet queries based on available features, such as contextual information, keywords and their semantic relationships. Beyond these methods, in this paper we propose to mine user interaction activities to predict the intent of the user during a navigation session. However, since in practice it is necessary to use a suitable mix of all such methods, it is important to exploit all the mentioned features in order to properly classify users based on their common intents. To this end, we have performed several experiments aiming to empirically derive a suitable classifier based on the mentioned features.