Some Eecient Solutions to the Aane Scheduling Problem Part Ii Multidimensional Time
نویسنده
چکیده
This paper extends the algorithms which were given in Part I to cases in which there is no aane schedule, i.e. to problems whose parallel complexity is polynomial but not linear. The natural generalization is to multi-dimensional schedules with lexicographic ordering as temporal succession. Multidimensional aane schedules, are, in a sense, equivalent to polynomial schedules, and are much easier to handle automatically. Furthermore, there is a strong connexion between multidimensional schedules and loop nests, which allows one to prove that a static control program always has a multidimensional schedule. Roughly, a larger dimension indicates less parallelism. In the algorithm which is presented here, this dimension is computed dynamically, and is just suucient for scheduling the source program. The algorithm lends itself to a \divide and conquer" strategy. The paper gives some experimental evidence for the applicability, performances and limitations of the algorithm. 1 2 R esum e Dans cet article, les algorithmes qui ont et e propos e dans la premi ere partie sont etendus au cas o u le programme source n'a pas de base de temps aane, c'est-a-dire a des algorithmes dont la complexit e parall ele est polynomiale mais non lin eaire. La solution naturelle est l'emploi de bases de temps a plusieurs dimensions, l'ordre de succession temporelle etant l'ordre lexi-cographique. Les bases de temps multidimensionnelles sont, en un certain sens, equivalentes a des bases de temps polynomiales, et sont beaucoup plus faciles a manipuler algorithmiquement. De plus, il y a une connexion forte entre bases de temps multidimensionnelles et nids de boucles, ce qui permet de d emontrer qu'un programme a contr^ ole statique a toujours une base de temps multidimensionnelle. En gros, plus grande est la dimension et moins il y a de parall elisme. Dans l'algorithme ici pr esent e, cette dimension est d etermin ee dynamiquement; elle est juste suusante pour permettre l'ordonnancement du programme source. Ennn, cet algorithme se pr^ ete a l'application de la strat egie \diviser pour r egner". On pr esente en conclusion quelques r esultats exp erimentaux permettant de juger du domaine d'application, des performances et des limitations de l'algorithme. 3 do i = 0,n do j = 0,i 1 s = s + a(i,j) end do end do Figure 1: A simple program with no linear schedule Instruction Domain 1 0 i n 0 j i
منابع مشابه
Aane-by-statement Scheduling of Uniform and Aane Loop Nests over Parametric Domains
This paper deals with parallel scheduling techniques for uniform and aane loop nests. We deal with aane-by-statement scheduling, a powerful extension of Lamport's hyperplane method where each statement within the loop nest is scheduled by a diierent timing function. We present a new, constructive and eecient method to determine the optimal (i.e. with smallest latency) aane-by-statement scheduli...
متن کاملA multi-objective resource-constrained optimization of time-cost trade-off problems in scheduling project
This paper presents a multi-objective resource-constrained project scheduling problem with positive and negative cash flows. The net present value (NPV) maximization and making span minimization are this study objectives. And since this problem is considered as complex optimization in NP-Hard context, we present a mathematical model for the given problem and solve three evolutionary algorithms;...
متن کاملA New Solution for the Cyclic Multiple-Part Type Three-Machine Robotic Cell Problem based on the Particle Swarm Meta-heuristic
In this paper, we develop a new mathematical model for a cyclic multiple-part type threemachine robotic cell problem. In this robotic cell a robot is used for material handling. The objective is finding a part sequence to minimize the cycle time (i.e.; maximize the throughput) with assumption of known robot movement. The developed model is based on Petri nets and provides a new method to calcul...
متن کاملSome Eecient Solutions to the Aane Scheduling Problem Part I One-dimensional Time
Programs and systems of recurrence equations may be represented as sets of actions which are to be executed subject to precedence constraints. In many cases, actions may be labelled by integral vectors in some iteration domain, and precedence constraints may be described by aane relations. A schedule for such a program is a function which assigns an execution date to each action. Knowledge of s...
متن کاملScheduling Uniform Loop Nests
This paper surveys scheduling techniques for uniform loop nests. First we introduce the hyperplane method and related variants. Then we extend it by using a diierent aane scheduling for each statement within the nest. In both cases we present a new, constructive and eecient method to determine optimal solutions.
متن کامل