Bodin, F, Beckman P, Gannon D, Narayana S, Yang SX. Distributed pC++: Basic ideas for an object parallel language. Scientific Programming 1993; 2:7--22.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....that needs to be sent and received based on the knowledge of data dependences. In contrast with fine grain loop level parallelism, we address coarse grain parallelism. Finally, there are many object models that add concurrency constructs to the C programming language to support data parallelism [8, 4, 9]. These models support HPFlike concurrency, in an object oriented setting. Instead, we address coarse grain concurrency. 7. Conclusions and Future Work This paper presented a distributed object model for data parallel irregular applications. The model distinguishes between distributed, active ....

F. Bodin, P. Beckman, D. Gannon, S. Narayana, and S. Yang. Distributed pc++: Basic ideas for an object parallel language, 1992.

....compiler technology is not yet advanced enough for fully automatic handling of data distributions, the user is required to supply additional information. In this context the use and development of new data parallel languages such as High Performance Fortran (HPF) 10] Vienna Fortran [6] and pC [3] gained considerable attention. For dedicated application domains portable libraries have been designed like ScaLAPACK [8] for linear algebra or PIPS [15] for image processing. The major aim of these systems is to ease the programming of parallel computers and to generate portable code. Since the ....

....seem to have become a kind of standard. The linaer algebra library ScaLAPACK [8] is a distributed memory version of the LAPACK software package for matrix computations. It utilizes subsets and combinations of data distribution. pC is an object oriented extension to the C programming language [3] . The data distributions allowed are identical to HPF s. 2.2 Data distributions for parallel image processing In this subsection we will describe the data distributions that we found to be of interest in the application field of parallel image processing. For low and medium level image ....

[Article contains additional citation context not shown here]

F. Bodin, P. Beckman, D. Gannon, S. Narayana, and S. X. Yang. Distributed pC++: Basic ideas for an object parallel language. Scientific Programming, 2(3), 1993.

....(to prototype complex computations) For instance, when coding e#ort is paramount and composition is implemented in a distributed objects system (e.g. 22, 39] techniques such as inheritance and templates can be used to create new components. Other implementations involving parallel programming [12, 19, 31] or multi agent coordination [25, 26] provide comparable facilities (typically APIs) for creating new components. Component composition e#ort, on the other hand, emphasizes the modeling of a computation as a process of, say, graphically laying out a network of components (e.g. 52] By providing ....

F. Bodin, P. Beckman, D. Gannon, S. Narayana, and S.X. Yang. Distributed pC++: Basic Ideas for an Object Parallel Language. Scientific Programming, Vol. 2(3):pages 7--22, 1993.

....prototype complex computations) For instance, when coding effort is paramount and composition is implemented in a distributed objects system 5 (e.g. 22, 39] techniques such as inheritance and templates can be used to create new components. Other implementations involving parallel programming [12, 19, 31] or multi agent coordination [25, 26] provide comparable facilities (typically APIs) for creating new components. Component composition effort, on the other hand, emphasizes the modeling of a computation as a process of, say, graphically laying out a network of components (e.g. 52] By ....

F. Bodin, P. Beckman, D. Gannon, S. Narayana, and S.X. Yang. Distributed pC++: Basic Ideas for an Object Parallel Language. Scientific Programming, Vol. 2(3):pages 7--22, 1993.

...., which has only the nonzero entries for row j of S. This representation of S is shown in Figure 2(b) Matrix N does not have to be represented explicitly. S = # # # # # #N = # # # 1 1 1 1 1 1 1 1 1 1 1 1 1 # # # #1 = J1 = 1 3 5 ] #2 = J2 = 1 2 5 ] #3 = J3 = [ 1 2 3 4 ] #4 = J4 = 2 3 4 ] a) b) Figure 2: Structure of the spending matrix S and normalizing matrix N (a) representation of matrix S (b) Let # i be the product dependent scaling factor for product i as discussed in the preceding section. The score # ij of customer j against product i is ....

....entries for row j of S. This representation of S is shown in Figure 2(b) Matrix N does not have to be represented explicitly. S = # # # # # #N = # # # 1 1 1 1 1 1 1 1 1 1 1 1 1 # # # #1 = J1 = 1 3 5 ] #2 = J2 = 1 2 5 ] #3 = J3 = 1 2 3 4 ] #4 = J4 = [ 2 3 4 ] (a) b) Figure 2: Structure of the spending matrix S and normalizing matrix N (a) representation of matrix S (b) Let # i be the product dependent scaling factor for product i as discussed in the preceding section. The score # ij of customer j against product i is computed as: # ij = # i # ....

[Article contains additional citation context not shown here]

F. Bodin, P. Beckmann, D. Gannon, S. Narayana, and S. X. Yang. Distributed pC++: Basic ideas for an object parallel language. Scientific Programming, 2(3):7--22, 1993.

....Both languages are most effective as SPMD programming environments. Neither language strongly supports the data parallel model as they require the programmer to explicitly perform synchronization and communication through dataflow like mechanisms. Both languages are extensions to C . pC [2] and C from the University of Wisconsin [11] both strongly support the data parallel model. However, both languages are extensions to C . pC also differs from our work in that they are building upon the High Performance Fortran instantiation of the data parallel programming model, whereas we ....

F. Bodin, P. Beckman, D. Gannon, S. Narayana, and S. Yang. Distributed pC++: Basic ideas for an object parallel language. In Object Oriented Numerics Conference, April 1993.

....skeletons, design patterns, and distributed objects. Our work is also in some respects complementary to work exploring the use of programming skeletons and patterns in 4 parallel computing, for example that of Cole [27] and Brinch Hansen [15] and even work exploring distributed objects, pC [12] for example. We also make use of abstractions that capture exploitable commonalities among programs, but we use these abstractions to guide a program development methodology based on program transformations. Communication libraries. Much work has been done in developing program libraries ....

....of structuring parallel programs in terms of dataflow; our work differs in that it addresses patterns of both dataflow and computation. 162 Distributed objects. The mesh spectral archetype is based to some extent on the idea of distributed objects, as discussed for example in work on pC [12] and POOMA [4] Our work on archetypes differs from this work in that we focus more on the pattern of computation and on identifying and exploiting patterns of computation and communication. Communication libraries. Many researchers have investigated and developed reusable general libraries of ....

F. Bodin, P. Beckman, D. Gannon, S. Narayana, and S. X. Yang. Distributed pC++: Basic ideas for an object parallel language. Scientific Programming, 2(3):7--22, 1993.

....limitations for block structured methods) and also tensor products of 1d block irregular decompositions. These mechanisms alone cannot describe the irregular blocking structures that arise in adaptive mesh refinement and recursive coordinate bisection [44] Data Parallel C Languages The pC [30, 31, 144] programming language is a data parallel extension of C . It implements a concurrent aggregate [47] model in which a parallel operation is applied simultaneously to all elements of a data aggregate called a collection. 4 Information about the second High Performance Fortran standardization ....

F. Bodin, P. Beckman, D. Gannon, S. Narayana, and S. X. Yang, Distributed pC++: Basic ideas for an object parallel language, Journal of Scientific Programming, 2 (1993).

....on shared memory architectures, appears in [36] Examples of specific algorithms include one deep mergesort [20] and one deep quicksort [13, 35] Distributed objects. The mesh spectral archetype is based to some extent on the idea of distributed objects, as discussed for example in work on pC [5] and POOMA [1] We differ from this work in that we focus more on the pattern of computation and on identifying and exploiting patterns of computation and communication. Communication libraries. Many researchers have investigated and developed reusable general libraries of communication routines; ....

F. Bodin, P. Beckman, D. Gannon, S. Narayana, and S. X. Yang. Distributed pC++: Basic ideas for an object parallel language. Scientific Programming, 2(3):7--22, 1993.

....limitations of traditional languages. Their major drawing power is to offer the opportunity to encapsulate complex coding and provide user friendly interface. Some attempts to exploit object oriented technology in parallel computing have recently been conducted (see for instance EPEE[6] pC [1] or LAPACK [5] One drawback for their propagation is the lack of standards (which OO language, which communication libraries, which data distribution) In this work, we question the benefits one could obtain in developing numerical libraries using a neat and widespread DOO language. We ....

F. Bodin, P. Beckman, D. Gannon, S. Narayana, S. Yang, Distributed pC++: basic ideas for an object parallel language. In Proceedings of the First Annual Object-Oriented Numerics Conference (OONSKI'93), Rogue Wave Software Publish., Sunriver, Oregon, 1993, pp. 1-24.

....distributions. Experimental results show that our software scheme not only can be easily implemented in the practice, but also is with good efficiency. 1. Introduction An increasing number of programming languages, such as Fortran D, HPF (High Performance Fortran) 3] and parallel C [13, 2], are providing distributed arrays to support a global name space on distributed memory architectures. The distribution pattern specified by programmers in HPF includes Block pattern which has each processor get a consecutive block of elements, Cyclic distribution which is # G.H. Hwang and ....

Bodin, F., Beckman, P., Gannon, D., Narayana, S., and Yang, S. Distributed pC++: Basic Ideas for an Object Parallel Language, Scientific Programming, 2(3), Fall 1993.

.... of A and S discussed previously, the scores for a customer j against all products in a S = 2 6 4 Theta Theta Theta Theta Theta Theta Theta Theta Theta Theta Theta Theta Theta 3 7 5 N = 2 6 4 1 1 1 1 1 1 1 1 1 1 1 1 1 3 7 5 oe 1 = Theta Theta Theta ] J1 = [ 1 3 5 ] oe 2 = Theta Theta Theta ] J2 = 1 2 5 ] oe 3 = Theta Theta Theta Theta ] J3 = 1 2 3 4 ] oe 4 = Theta Theta Theta ] J4 = 2 3 4 ] Figure 2: Structures of the spending matrix S and normalizing matrix N , and representation of matrix S. 3 product group i can be computed in ....

.... for a customer j against all products in a S = 2 6 4 Theta Theta Theta Theta Theta Theta Theta Theta Theta Theta Theta Theta Theta 3 7 5 N = 2 6 4 1 1 1 1 1 1 1 1 1 1 1 1 1 3 7 5 oe 1 = Theta Theta Theta ] J1 = 1 3 5 ] oe 2 = Theta Theta Theta ] J2 = [ 1 2 5 ] oe 3 = Theta Theta Theta Theta ] J3 = 1 2 3 4 ] oe 4 = Theta Theta Theta ] J4 = 2 3 4 ] Figure 2: Structures of the spending matrix S and normalizing matrix N , and representation of matrix S. 3 product group i can be computed in the following way. First, two vectors and j ....

[Article contains additional citation context not shown here]

F. Bodin, P. Beckmann, D. Gannon, S. Narayana, and S. X. Yang. Distributed pC++: Basic ideas for an object parallel language. Scientific Programming, 2(3):7--22, 1993.

....extensions have appeared. The most popular are Fortran D[8] Fortran M[6] HPF. Looking at C language it gets really harder to design an efficient parallel extension. Few extensions already have been designed, all based on different concepts. Some of them have emerged, namely CC [3] pC [1] or C [2] More formally, a US consortium called High Performance C (HPC 1 ) also tries to define a standard model for parallel programming based on C . This model is aimed at enabling the writing of portable parallel applications and at providing a target language for vendors allowing ....

F. Bodin, P. Beckman, D. Gannon, S. Naranaya, and S. X. Yang. Distributed pC++: Basic Ideas for an Object Parallel Language. Scientific Programming, 2(3), 1993.

....and detect parallelizable loops. Their work illustrates how semantic information about standard classes exposes new optimization and parallelization opportunities. Array operations are natural candidates for exploitation of parallelism. This has been done extensively before both at the language [11, 34, 8] and subsystem [18, 28, 26] levels, with different degrees of transparency to the application programmer. So far, we have successfully exploited parallelism inside the Array package operations while achieving total transparency to the programmer. A combination of the Array package with semantic ....

F. Bodin, P. Beckmann, D. Gannon, S. Narayana, and S. X. Yang. Distributed pC++: Basic ideas for an object parallel language. Scientific Programming, 2(3):7--22, 1993.

....by means of a graphical interface and an API; to visualize arrays using tools such as IRIS Explorer. This implies to setup definitions: definition of services; definition of how data can be accessed; definition of a protocol to dialogue with between tools. A prototype implementation, based on PC [12] with visualization achieved by means of IRIS Explorer, has been around for some time. An HPF implementation has recently been released. An other very interesting project is EPPP[36] It offers a development platform, ranging from the language to a performance debugger. The language underlying the ....

F. Bodin, P. Beckman, D. Gannon, S. Narayana, and S.X. Yang. Distributed pC++: Basic Ideas for an Object Parallel Language. Scientific Programming, 2(3):61--74, 1993.

....block decompositions [19] for regular applications. DOCK includes two classes, Processor and Decomposition, that implement first class C versions of the virtual processor arrays and templates in High Performance Fortran [20] The DOCK classes are modeled after similar classes introduced in pC [7]. Using the DOCK mapping classes, we first declare a virtual processor array. By default, the Processors2 object creates a 2D logical processor array corresponding to the physical processor set (2) Next, we must decompose the global domain across the virtual processor set. The Decomposition ....

....dynamic structured applications [24] Structural abstraction derives from the domain abstractions used in the FIDIL programming language [21] LPARX to appear in J. Parallel and Distributed Computing uses an asynchronous communication model, as do several other parallel languages and libraries [15, 23, 7, 11, 10]. An asynchronous communication model is most appropriate for applications where run time preprocessing of the communication pattern is impossible. However, when communication dependences can be computed beforehand, run time preprocessing can significantly reduce communication overheads and ....

[Article contains additional citation context not shown here]

Bodin, F., Beckman, P., Gannon, D., Botwals, J., Narayana, S., Srinivas, S., and Winnicka, B. Distributed pC++: Basic ideas for an object parallel language. Journal of Scientific Programming 2, 3 (1993), 7-22.

....are restricted to tensor products of 1d irregular decompositions, and it is unclear how Vienna Fortran will support dynamic irregular blocking structures such as those required by adaptive mesh refinement and recursive coordinate bisection decompositions. The pC programming language [10] implements a collection abstraction which includes a coarse grain data parallel loop over objects within the collection; pC employs a data decomposition scheme similar to that of HPF. The Multiblock PARTI [1] and CHAOS [16] libraries provide run time support for data parallel compilers such as ....

F. Bodin, P. Beckman, D. Gannon, S. Narayana, and S. X. Yang, Distributed pC++: Basic ideas for an object parallel language, Journal of Scientific Programming, 2 (1993).

....Groups Architecture SIG, Applications SIG, and Implementations SIG. Between these three groups, the aim is to identify emerging de facto standards for Parallel C compilers and ultimately produce a standard definition and library architecture for parallel C compilers. 5.2. 7 pC Sage pC [21] is a portable parallel C for high performance computers. pC is a language extension to C that permits data parallel style operations using collections of objects from some base element class. Member functions from this element class can be applied to the entire collection in parallel. ....

F. Bodin, P. Beckman, D. B. Gannon, S. Narayana, and S. Yang. Distributed pC++: Basic ideas for an object parallel language. In Proc. Supercomputing '91, pages 273--282, 1991.

....parallel functions and operators to aggregate data structures. In an object parallel system, the programmer creates and manipulates collection or container objects that encapsulate a distributed data structure on a massively parallel computer. Examples of this style of programming include 1. pC [1], a collection based parallel programming language. 2. Single Program Multiple Data (SPMD) C class libraries that provide user defined, encapsulated support for parallel, distributed data structures. Examples include LPARX[2] P [3] the POOMA library[4] and CHAOS [5] 3. Data parallel ....

D. Gannon, F. Bodin, P. Beckman, S. Yang, and S. Narayana. Distributed pC++: Basic ideas for an object parallel language, Journal of Scientific Programming, Vol. 2, pp. 7-22 (1993).

....and Charm , based on message driven objects, have been retargeted on top of Converse. Prototype implementations of SPMD languages such as SM, PVM and MPI [1] have been done using Converse framework. A data parallel language DP [11] subset of HPF) and a data parallel object oriented language pC [2] have been implemented on top of Converse. A parallel simulation language, Import [12] and a functional language Agents [15] are under development. Converse has also been used to enrich scripting languages such as Perl to support message driven objects. Message driven extensions to Java [7] are ....

F. Bodin, P. Beckman, D. Gannon, S. Narayana, and S. Yang. Distributed pC++: Basic Ideas for an Object Parallel Language. Scientific Programming, 2(3), 1993.

No context found.

Bodin, F, Beckman P, Gannon D, Narayana S, Yang SX. Distributed pC++: Basic ideas for an object parallel language. Scientific Programming 1993; 2:7--22.

No context found.

Bodin, F., Beckman, P., Gannon, D., Narayana, S., and Yang, S. Distributed pC++: Basic Ideas for an Object Parallel Language, Scientific Programming, 2(3), Fall 1993.

No context found.

F. Bodin, P. Beckman, D. Gannon, S. Narayana, and S. Yang. Distributed pC++: Basic Ideas for an Object Parallel Language. In Proceedings of the First Annual ObjectOriented Numerics Conference, pages 1--24, April 1993. Efficient Compilation of Out-of-core Data Parallel Programs

No context found.

Bodin, F., Beckman, P., Gannon, D., Narayana, S., and Yang, S. Distributed pC++: Basic Ideas for an Object Parallel Language, Scientific Programming, 2(3), Fall 1993.

No context found.

F. Bodin, P. Beckman, D. Gannon, S. Narayana, and S. Yang. Distributed pC++: Basic Ideas for an Object Parallel Language. In Proceedings of the First Annual Object-Oriented Numerics Conference, pages 1--24, April 1993.

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