D. DeGroot. Restricted AND-Parallelism. In International Conference on Fifth Generation Computer Systems, pages 471--478. Tokyo, November 1984.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....program concurrent execution. The other direction aims at the parallel execution of pure logic or Prolog programs that involve searching. Programmers may specify control only for guiding execution. OR parallelism was exploited first [2] 26] 8] 43] and AND parallelism has been incorporated also [9][16] 45] 15] There has been a long history of controversy between the proponents of these two directions. Proponents of concurrent logic programming languages claim that pure logic programs are not expressive enough to describe e#cient parallel algorithms or to make e#ective use of the ....

DeGroot, D. Restricted AND-Parallelism. In Proc. Int. Conf. on FGCS'84, ICOT, 1984, pp. 471--478.

....a user language of GHC for search problems. Exploiting parallelism in an obtained GHC program is quite easy, because di erent search paths are examined by independent ANDparallel goals without communication. This means that an obtained program could be processed using restricted AND parallelism [5] also. A deterministic program obtained by the previous method simulates OR parallel and AND sequential execution of the original program. The original OR parallelism is compiled into AND parallelism as stated above, and the sequential execution of conjunctive goals is 3 realized by passing a ....

....in Program 2 instantiates X to [1,2 X2] nocheck(X) must be invoked to detect that [1,2 X2] cannot be a solution for any instance of X2. At the same time, we must compile away the bagof primitive that is outside Horn clause logic. 4 : bagof(X, eightqueens(X) B) eightqueens(X) perm([1,2,3,4,5,6,7,8],X) nocheck(X) perm( H T] A P] del( H T] A,L) perm(L,P) perm( del( H T] H, T ) del( H T] A, H T2] del(T,A,T2) nocheck( H T] qsafe(H,T,1) nocheck(T) nocheck( qsafe(U, H T] N) H N= U, H N= U, M is N 1, qsafe(U,T,M) qsafe( Program ....

[Article contains additional citation context not shown here]

DeGroot, D., Restricted AND-Parallelism. In Proc. Int. Conf. on Fifth Generation Computer Systems

....be profitable to investigate evaluation strategies where items associated with ground terms are evaluated first. For example, it might be useful to use dependency relations of AV grammars to find a ground first evaluation order. Techniques for keeping track of groundness properties can be found in [DeG84]. Such evaluation strategies may even be profitable for sequential implementations. Grammars It is possible that the amount of parallelism that can be exploited is for a great part dependent on the type of grammar that is used. It might be possible to automatically rewrite a grammar so that ....

Doug DeGroot. Restricted and-parallelism. In ICOT, editor, Proceedings of the International Conference on Fifth Generation Computer Systems (FGCS '84), pages 471--478, November 1984. 26

....be profitable to investigate evaluation strategies where items associated with ground terms are evaluated first. For example, it might be useful to use dependency relations of AV grammars to find a ground first evaluation order. Techniques for keeping track of groundness properties can be found in [DeG84]. Such evaluation strategies may even be profitable for sequential implementations. Grammars It is possible that the amount of parallelism that can be exploited is for a great part dependent on the type of grammar that is used. It might be possible to automatically rewrite a grammar so that ....

Doug DeGroot. Restricted and-parallelism. In ICOT, editor, Proceedings of the International Conference on Fifth Generation Computer Systems (FGCS '84), pages 471--478, November 1984.

....allowed to be executed concurrently by different andagents (and agents are processing agents working in and parallel with each other) Dependencies are detected at run time by executing some simple tests in troduced by the parallelizing compiler. ACE adopts the technique originally designed by DeGroot (DeGroot 1984) and refined by Hermenegildo (Hermenegildo et al. 1995) of annotating the program at compile time with Conditional Graph Expressions (CGEs) h conditions i ) B 1 Delta Delta Delta Bn ) where h conditions i is a conjunction of simple tests on variables appearing in the clause that verifies ....

DeGroot, D. 1984. Restricted AND-Parallelism. In Conf. on 5th Generation Computer Systems.

....pre condition for ensuring the correctness and usefulness of many important optimizations. This is exemplified in program parallelization where different notions of independence [9] and the related concept of stability [7] are the basis of models which incorporate Independent And parallelism [3, 5, 9, 12, 20] as in these models the parallel execution of a set of goals in the body of a clause is ensured to be correct and efficient w.r.t. the sequential execution if the goals are proved to be independent. Independence is also the basis of optimizations such as intelligent backtracking [16] and goal ....

....it takes to compute such solutions. This time can change either because the actual number of reduction steps differs and or because the amount of work involved in performing each of those steps differs in a significant way. Previous work in the context of traditional Logic Programming languages [3, 5, 9] has concentrated on defining independence in terms of preservation of search space. This has been achieved by ensuring that either the goals do not share variables (strict independence) or if they share variables, that they do not compete for their bindings (non strict independence) It could ....

[Article contains additional citation context not shown here]

D. DeGroot. Restricted AND-Parallelism. In International Conference on Fifth Generation Computer Systems, pages 471--478. Tokyo, November 1984.

....on the results obtained, it is concluded that the overhead of global flow analysis is not prohibitive, while the results of analysis can be quite precise and useful. 1 Introduction The extensive use of advanced compilation techniques [8, 22, 30, 32, 33, 34] coupled with parallel execution [5, 10, 15, 20, 35], appears to be a very promising approach to achieving improved performance in logic programming systems. Existing systems are based largely on local analysis (i.e. clause level or, at most, procedure level, as in the WAM) Such techniques have already brought substantial performance improvements ....

....be expected in this case, the inability to represent and reason about free variables results in lower absolute precision figures. 4 An Application: And parallelism Detection This section discusses the application of mode inferencing to the generation of Independent Restricted Andparallelism [10, 15, 14], an efficient type of parallelism in which only independent goals are executed in parallel and one of the main applications of the MA 3 system. Note, however, that the application of mode information is in general much broader, ranging from other high level applications, such as the improvement ....

[Article contains additional citation context not shown here]

D. DeGroot. Restricted AND-Parallelism. In International Conference on Fifth Generation Computer Systems, pages 471--478. Tokyo, November 1984.

....Our objective is to apply the methodology sketched in the preceding section and develop paradigms for the visualization of parallel execution models for logic programs. We will focus on the visualization of Or parallelism [Lus88, CSW88, CA88, AK90, Kar92] Restricted And parallelism (RAP) DeG84, HG90] and Determinate Dependent And parallelism (DDAP) as in [Nai88, SCWY90] and committed choice languages. We are also interested in visualizing combinations of these models. 3.1 Common Design Concepts Examples of common observables, aimed at studying the systems of interest under the ....

D. DeGroot. Restricted AND-Parallelism. In International Conference on Fifth Generation Computer Systems, pages 471--478. Tokyo, November 1984.

....other execution orders, or run parts of the program in parallel, as long as no violation of the language s operational semantics can be observed. The simplest case is, of course, when there are no data dependencies between the various instances of the body (we may say that they are independent 9 [16]) In this case, all operational semantics above are equivalent and the implementation may choose to run them in any order or all at once. When bounded quantifications are added to a logic programming language, their operational semantics should follow that of the existing language constructs, ....

DeGroot, D., Restricted AND-Parallelism, in: Proc. Intl. Conf. on Fifth Generation Computer Systems 1984, North-Holland, Amsterdam, 1984.

....backtracking in the presence of shared variables. The nave method of attacking this problem would be to avoid sharing altogether, having each goal compute its own solutions independently and combining them afterward to find the complete solutions. This method is seen to be infeasible [Conery 1987, DeGroot 1984] thus, some other approach must be taken. One approach that taken by the independent AND parallel systems achieves parallelism by running independent goals concurrently. Dependent goals that is, goals that share variables are run sequentially, usually using some form of backtracking to ....

....in parallel. Conery s AND OR process model [Conery 1987] uses dynamic analysis and needs no annotations. Because goal independence is determined at run time, independent parallelism is maximized, though at the cost of a high run time overhead. DeGroot s Restricted AND Parallelism (RAP) model [DeGroot 1984] takes a hybrid approach, combining annotations generated at compile time with simple run time checks for independence to handle difficult cases. The concurrent logic programming (CLP) languages take the opposite approach, trying to maximize parallelism by allowing all goals including those ....

Doug DeGroot. Restricted AND-parallelism, In Proceedings of the 1984 International Conference on Fifth Generation Computer Systems, pages 471--478, Tokyo, November 1984.

....The existing efficient parallel Prolog systems exploit only one form of parallelism, namely either independent AND parallelism (as in [50] or OR parallelism (as in [65] and Chapter 3) This thesis is concerned with the latter. The independent AND parallelism approach was first taken by DeGroot [32], and then by Hermenegildo, who designed RAP WAM, an extension of the WAM for independent AND parallelism [50] An alternative independent AND parallel system has been developed by Lin and Kumar [61] Data dependency analysis and compilation for parallel execution are important themes for ....

....to be tried in parallel, while AND parallelism allows the goals of a clause body to be executed in parallel. There are two forms of AND parallelism: independent AND parallelism, and stream AND parallelism. In independent ANDparallelism goals that do not share any variable are executed in parallel [32, 50]. Stream AND parallelism allows goals that share variables to be executed in parallel with the value of the shared variable being communicated incrementally between the goals [36, 78, 86] In Prolog, OR parallelism and independent AND parallelism are much easier to implement efficiently than ....

[Article contains additional citation context not shown here]

Doug DeGroot. Restricted AND-Parallelism. In Hideo Aiso, editor, International Conference on Fifth Generation Computer Systems 1984, pages 471--478. Institute for New Generation Computing, Tokyo, 1984.

.... Prolog [38] encapsulates the dependency information in the code generated by the compiler. In this way simple and inexpensive runtime checks can be done to identify sources of dependency. This technique was first devised by DeGroot and is called Restricted (or Fork Join) And Parallelism (RAP) [17], and was further enhanced by Hermenegildo [42, 43, 45] It does not capture all the and parallelism, although it does manage to exploit a substantial portion of it (see later) Moreover, research has shown that generation of these annotations through compile time analysis [15, 24, 41, 7, 46] ....

....arbitrary and parallelism (i.e. not only independent) The same language is used to make the result of the automatic parallelization visible to the user if so desired. In this system, program clauses are compiled into Conditional Graph Expressions (CGEs) originally developed by DeGroot [17] and successively refined by Hermenegildo [43] and Shen [70] Conditional Graph Expressions are expressions of the form (condition ) goal 1 goal 2 : goal n ) meaning that, if condition is true, goals goal 1 : goal n should be evaluated in parallel, otherwise they should be ....

[Article contains additional citation context not shown here]

D. DeGroot. Restricted AND-Parallelism. In Int'l Conf. on 5th Generation Computer Systems, pages 471-- 478. Tokyo, Nov. 1984.

....(named ACE running on a Sequent Symmetry Multiprocessor) to illustrate LPCO and NPCO. 2 Independent And parallelism In this section we briefly describe independent andparallelism and how it is implemented. Foundations of much of the work described in this section were laid down in [6, 3], however, the specific implementation described is that of the ACE system [9] Conventionally, an and parallel Prolog system works by executing a program that has been annotated with parallel conjunctions. These parallel conjunction annotations are either inserted by a parallelizing compiler ....

....Also, for many programs the number of parallel conjunctions that can be combined into one will only be determined at run time. For example, consider the following program: processlist( H T] Hout Tout] process(H, Hout) processlist(T, Tout) processlist( processlist([1,2,3,4], Out) In such a case, compile time transformations cannot unfold the program to eliminate nesting of parcall frames because it will depend on the length of the input list. However, using our runtime technique, since the the goal process list is determinate, nesting of parcall frames can be ....

[Article contains additional citation context not shown here]

D. DeGroot. Restricted AND-parallelism. In International Conference on Fifth Generation Computer Systems, Nov., 1984.

....to right traversal of the AND OR tree built during program execution. Parallelism in logic programs is implicit in the sense that no extra annotation is required. Parallel implementations are concentrated either in the exploitation of ANDparallelism (a conjunction of goals is resolved in parallel) [6,7,8] or OR Parallelism (a goal is resolved by many different ways) 1,5,11,12] However, many approaches, such as the Committed Choice Non Deterministic languages [4,10,14,16] sacrifice the full semantics of Prolog for the sake of efficiency. Lately, there has been an attempt to exploit both types of ....

....many different ways) 1,5,11,12] However, many approaches, such as the Committed Choice Non Deterministic languages [4,10,14,16] sacrifice the full semantics of Prolog for the sake of efficiency. Lately, there has been an attempt to exploit both types of parallelism within the same implementation [3,7,20]. Performance comparisons between models which exploit only AND parallelism and models which exploit only OR parallelism suggested that a high performance system should have backtracking, full unification and dependent AND Parallelism [15] This paper looks at execution of logic programs from a ....

D.DeGroot, Restricted AND-parallelism, FGCS' 84 Conference, Tokyo (1984) 471478.

....concurrently by different andagents (and agents are processing agents working in andparallel with each other) Dependencies are detected at run time by executing some simple tests introduced by the parallelizing compiler. In ACE we have adopted the technique originally designed by DeGroot [3] and refined by Hermenegildo [7] adopted also by Prolog [6] of annotating the program at compile time with Conditional Graph Expressions (CGEs) This will be explained in detail in a later section. Since and agents are computing just different parts of the same computation (i.e. they are ....

D. DeGroot. Restricted AND-Parallelism. In Int'l Conf. on 5th Generation Computer Systems, pages 471--478. Tokyo, Nov. 1984.

....are allowed to be executed concurrently by different and agents (and agents are processing agents working in and parallel with each other) Dependencies are detected at run time by executing some simple tests introduced by the parallelizing compiler. ACE adopts the technique originally designed by DeGroot (DeGroot 1984) and refined by Hermenegildo (Hermenegildo Greene 1990) of annotating the program at compile time with Conditional Graph Expressions (CGEs) h conditions i ) B 1 Delta Delta Delta Bn ) where h conditions i is a conjunction of simple tests on variables appearing in the clause that ....

DeGroot, D. 1984. Restricted AND-Parallelism. In Conf. on 5th Generation Computer Systems.

....h2; i is reached (wher 2 denotes an empty resolvent) When a failure occurs, backtracking (backward execution) is performed, which involves choosing an untried alternative in the subgoals to the left of the failing one. The sequential computation model above can be extended to exploit IAP [5, 14]. Subgoals are considered independent if the bindings produced by one of them do not a ect the computation of the others 2 . Formally, two subgoals p; q are said to be independent i 8R; n (hp; i n P hR; i 8R 0 ; m (hq; i m P hR 0 ; i hq ; i m P hR 0 ; ....

D. DeGroot. Restricted AND-Parallelism. In Int'l Conf. on 5th Generation Computer Systems, pages 471-478. Tokyo, Nov. 1984.

....(as in Prolog) This is in contrast with some of the previous works that were based on the notion of goal reuse [6, 16] goal reuse, in fact, has been shown to create situations where Prolog semantics cannot be guaranteed. For simplicity and efficiency, we adopt the solution proposed by DeGroot [9] of restricting parallelism to a nested parbegin parend structure. This is illustrated in Figure 1, which sketches the structure of the computation tree created in presence of an and parallel computation. The different branches are assigned to different and agents (and agents are processors ....

....access to terms created in the stacks of different and agents. Since we are exploiting independent and parallelism, only independent subgoals are allowed to be executed concurrently by different and agents. In order to ensure this, in ACE we have adopted the proposal originally designed by DeGroot [9] and refined by various researchers [14, 22, 27] of annotating the program at compile time with Conditional Graph Expressions (CGEs) A conditional graph expression (CGE for simplicity) is an expression of the form: h conditions i ) B 1 Delta Delta Delta Bn ) where h conditions i is a ....

[Article contains additional citation context not shown here]

DeGroot, D. Restricted AND-Parallelism. In Int'l Conf. on 5th Generation Computer Systems (1984).

....support for Prolog semantics) A DDAS program is essentially a Prolog program containing parallel annotations. The parallel annotations are expressed using Extended Conditional Graph Expressions (ECGE) 26] which represent a generalization of the concept of CGE, originally designed by DeGroot [11] and refined and improved by Hermenegildo [15] The purpose of an ECGE is twofold: 1. identify a sequence of subgoals that are meant to be executed (conditionally or unconditionally) in parallel; 2. identify the variables that are common to different parallel subgoals (shared or dependent ....

D. DeGroot. Restricted AND-Parallelism. In Int'l Conf. on 5th Generation Computer Systems, Tokyo, Nov. 1984.

....(named ACE running on a Sequent Symmetry Multiprocessor) to illustrate LPCO and NPCO. 2 Independent And parallelism In this section we briefly describe independent andparallelism and how it is implemented. Foundations of much of the work described in this section were laid down in [3], however, the specific implementation described is that of the ACE system [9] Conventionally, an and parallel Prolog system works by executing a program that has been annotated with parallel conjunctions. These parallel conjunction annotations are either inserted by a parallelizing compiler ....

....implemented quite efficiently by the authors in the ACE system [9] ACE implementation itself is inspired by the RAPWAM [7] 3 . ACE is an extension to the sequential WAM (Warren Abstract Machine [14] for and parallel execution of Prolog programs with and parallel annotation (such as CGEs [3]) The ACE system has shown remarkable results on a variety of benchmarks. Its performance figures for the Sequent Symmetry multiprocessor can be found in [9] 3 Other implementations based on the principles of RAPWAM have also been proposed in the past, like Prolog [7] and DDAS [12] 3 3 ....

[Article contains additional citation context not shown here]

D. DeGroot. Restricted AND-parallelism. In International Conference on Fifth Generation Computer Systems, Nov., 1984.

No context found.

D. DeGroot. Restricted AND-Parallelism. In International Conference on Fifth Generation Computer Systems, pages 471--478. Tokyo, November 1984.

No context found.

D. DeGroot. Restricted AND-Parallelism. In International Conference on Fifth Generation Computer Systems, pages 471--478. Tokyo, November 1984.

No context found.

D. DeGroot. Restricted AND-Parallelism. In International Conference on Fifth Generation Computer Systems, pages 471--478. Tokyo, November 1984.

No context found.

D. DeGroot. Restricted AND-Parallelism. In Int. Conf. on Fifth Generation Computer Systems, pages 471--478. Tokyo, November 1984.

No context found.

D. DeGroot. Restricted AND-Parallelism. In International Conference on Fifth Generation Computer Systems, pages 471--478. Tokyo, November 1984.

First 50 documents  Next 50

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC