G. A. Geist and Vaidy Sunderam. The Evolution of the PVM Concurrent Computing System. In COMPCON, February 1993.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....in River. 4.4. 2 Programming environments There are a number of popular parallel programming environments that support the single program, multiple data (SPMD) programming style, including messaging passing environments such as Message Passing Interface [The93] and Parallel Virtual Machine [GS93] as well as explicit parallel languages, such as Split C [CDG 93] These packages all provide a simple model of parallelism to the user, thus allowing the ready construction of parallel applications. However, none provide any facility to avoid run time perturbations or adapt to hardware ....

Al Geist and Vaidy Sunderam. The Evolution of the PVM Concurrent Computing System. In COMPCON, pages 549--557, February 1993.

....of idle workstations, and the ability to give them up immediately when the owner returns 32 . Meeting these requirements is the main feature that distinguished the above systems from other systems that allow users to obliviously spawn off processes to multiple hosts, such as PVM 33 [560, 229]. In technical terms, these requirements translate into monitoring facilities and some way to cope with the changing availability of workstations. The two main schemes that are used are process migration and using malleable jobs. Process migration has the advantage of being transparent to users, ....

G. A. Geist and V. S. Sunderam, "The evolution of the PVM concurrent computing system". In 38th IEEE Comput. Soc. Intl. Conf. (COMPCON), pp. 549--557, Feb 1993.

....identification of idle workstations, and the ability to give them up immediately when the owner returns. Meeting these requirements is the main feature that distinguished the above systems from other systems that allow users to obliviously spawn off processes to multi63 ple hosts, such as PVM 19 [333, 135]. In technical terms, these requirements translate into monitoring facilities and process migration facilities. Naturally, the priority given to the workstations owners implies interference with the performance of the parallel programs being run in the background [191, 215] The performance ....

G. A. Geist and V. S. Sunderam, "The evolution of the PVM concurrent computing system". In 38th IEEE Comput. Soc. Intl. Conf. (COMPCON), pp. 549--557, Feb 1993.

....tightly coupled machines. Over the years, a number of parallel programming languages and run time environments have been developed for distributed systems. Some of the earliest programming environments focus on coarse grain parallel applications; for example, p4 [108] Data parallel C [127] PVM [63], PARFORM [28] and MPI [161] Later systems use objects which can be migrated between idle nodes and invoked regardless of location; for example, Emerald [84] and CHARM [146] Finally, the most sophisticated systems, developed more recently, adjust the level of parallelism in the application to ....

....process. ffl Synchronization: messages synchronizing a set of cooperating processes. A participating process cannot continue until it has been notified that all cooperating processes have reached the synchronization point. In message passing programs, such as those using MPI [161] or PVM [63], synchronization occurs between pairs of processes performing sends and receives. However, for simplicity in our analysis, we consider only synchronization operations across all processes within a job (e.g. a barrier) As initially discussed in Section 3.3.1, when a communicating process must ....

G. A. Geist and Vaidy Sunderam. The Evolution of the PVM Concurrent Computing System. In COMPCON, February 1993.

....and Vitesse. Closely coupled MPP systems are built by Convex. As to SCI connected distributed systems, there is a number of companies implementing hardware connections. The low cost is an important factor in the advance of the parallel processing. The emerging software platforms such as PVM [11] and Linda [12, 13] in the last few years are strong evidence. With the current network latency and bandwidth, solving a problem in parallel on a number of workstations is not very effective. The cost effectiveness of using existing workstations is so appealing that more and more applications are ....

G. Geist and V.S. Sunderam. Evolution of the PVM concurrent computing system. In 1993 IEEE Compcon, pages 549--557, Feb. 1993.

....load on line (at run time) However, longterm adaptation could also be useful in our system. 5. 2 Programming Environments There are a number of popular parallel programming environments that support the SPMD programming style, including messaging passing environmentssuch as MPI [45] and PVM [20], as well as explicit parallel languages, such as Split C [17] These packages all provide a simple model of parallelism to the user, thus allowing the ready construction of parallel applications. However, none provide any facility to avoid run time perturbations or adapt to hardware devices of ....

G. Geist and V. Sunderam. The Evolution of the PVM Concurrent Computing System. In COMPCON, February 1993.

....is suitable for parallelization, this paper investigates how the sequential code can be parallelized while minimizing the modifications to the initial code and hardware investment. In this spirit, the parallel multigroup solution of the transport equation using PVM (Parallel Virtual Machine) [1,2], a public domain software package for parallelization on a network of workstations, is presented. Parallelization over the energy groups as well as other iterative scheme are examined. Section II describes the multigroup neutron transport equations. In Section III, the main steps for the ....

G.A. Geist and V.S. Sunderam, The Evolution of the PVM Concurrent Computing System, Proceeding -- 26th IEEE Compcon Symposium, pp. 471-478, San Francisco, February 1993.

....tightly coupled machines. Over the years, a number of parallel programming languages and run time environments have been developed for distributed systems. Some of the earliest programming environments focus on coarse grain parallel applications; for example, p4 [104] Data parallel C [122] PVM [57], PARFORM [25] and MPI [155] Later systems use objects which can be migrated between idle nodes and invoked regardless of location; for example, Emerald [78] and CHARM [141] Finally, the most sophisticated systems, developed more recently, adjust the level of parallelism in the application to ....

....our analysis ignores this intervening work under the assumption that the amount of work performed is small, applications can still use split phase operations with implicit coscheduling. 46 have reached the synchronization point. In message passing programs, such as those using MPI [155] or PVM [57], synchronization occurs between pairs of processes performing sends and receives. However, for simplicity in our analysis, we consider only synchronization operations across all processes within a job (e.g. a barrier) As initially discussed in Section 3.3.1, when a communicating process must ....

G. A. Geist and Vaidy Sunderam. The Evolution of the PVM Concurrent Computing System. In COMPCON, February 1993.

....space of large loops and pre fetches pages that will be needed by the loop. 1. 1 Comparisons with related works There are numerous works which deal with efficient implementations of parallel programming languages using the operating system approach, namely multi threaded packages like PVM [13], MAXI [12] and Linda [14] Compilers for parallel languages are rarer, used mostly for functional data flow languages (e.g. 28] where the natural parallelism and the challenge of implementing Strictness analysis and Lazy evaluation motivated most of the works. Several abstract parallel ....

G. A. Geist and V. S. Sunderam. The evolution of the PVM concurrent computing system. In IEEE Comput. Soc. Intl. Conf. (COMPCON), number 38, pages 549--557, Feb 1993.

....of edges. O(jEj) work for All Pair Shortest Path implies an efficient solution for transitive closure, which is known to be hard [3] X[i] X[f(i) X[h(i) 2 3 0 1 2 3 4 5 6 7 X[0] 1; X[1] 3; X[2] X[0] X[1] X[3] X[0] X[1] X[4] X[0] X[1] X[5] X[2] X[3] X[6] = X[3] X[4] X[7] X[5] X[6] for i= 2 . 7 do Figure 2: Relation between algebraic circuit and IR. Note that in order for algebraic circuits to be modeled by IR loops they must satisfy that the number of edges is O(n) This is indeed the case when the algebraic operators are binary or ....

.... Shortest Path implies an efficient solution for transitive closure, which is known to be hard [3] X[i] X[f(i) X[h(i) 2 3 0 1 2 3 4 5 6 7 X[0] 1; X[1] 3; X[2] X[0] X[1] X[3] X[0] X[1] X[4] X[0] X[1] X[5] X[2] X[3] X[6] X[3] X[4] X[7] X[5] X[6]; for i= 2 . 7 do Figure 2: Relation between algebraic circuit and IR. Note that in order for algebraic circuits to be modeled by IR loops they must satisfy that the number of edges is O(n) This is indeed the case when the algebraic operators are binary or unary. The relation between algebraic ....

[Article contains additional citation context not shown here]

G. A. Geist and V. S. Sunderam. The evolution of the PVM concurrent computing system. IEEE Comput. Soc. Intl. Conf. (COMPCON), (38):549--557, February 1993.

No context found.

G. A. Geist and V. S. Sunderam, "The Evolution of the PVM Concurrent Computing System", Proceedings -- 26th IEEE Compcon Symposium, pp. 471-478, San Fransisco, February 1993.

....of some of the more interesting and important facets of the PVM system, including the design philosophy, computing model, performance issues, and application experiences. Detailed expositions as well as pedagogical material on various aspects of the PVM system may be found, for example, in [12, 13, 14], in addition to the papers already cited. At the time of writing, PVM continues to be a popular and widely used system for concurrent computing, and it is expected that the project will mature and evolve even further in the future. In this section we briefly discuss ongoing auxiliary projects, ....

G. A. Geist and V. S. Sunderam, "The Evolution of the PVM Concurrent Computing System", Proceedings -- 26th IEEE Compcon Symposium, pp. 471-478, San Fransisco, February 1993.

....of some of the more interesting and important facets of the PVM system, including the design philosophy, computing model, performance issues, and application experiences. Detailed expositions as well as pedagogical material on various aspects of the PVM system may be found, for example, in [12, 13, 14], in addition to the papers already cited. At the time of writing, PVM continues to be a popular and widely used system for concurrent computing, and it is expected that the project will mature and evolve even further in the future. In this section we briefly discuss ongoing auxiliary projects, ....

G. A. Geist and V. S. Sunderam, "The Evolution of the PVM Concurrent Computing System", Proceedings -- 26th IEEE Compcon Symposium, pp. 471-478, San Fransisco, February 1993.

No context found.

G. A. Geist and V. S. Sunderam, "The Evolution of the PVM Concurrent Computing System", Proceedings -- 26th IEEE Compcon Symposium, pp. 471-478, San Francisco, February 1993.

No context found.

G. A. Geist and Vaidy Sunderam. The Evolution of the PVM Concurrent Computing System. In COMPCON, February 1993.

No context found.

G.A. Geist and V.S. Sunderam, The Evolution of the PVM Concurrent Computing System, Proceedings - 26th IEEE Compcon Symposium, pp. 471-478, San Fransisco, February 1993.

No context found.

G.A. Geist and V.S. Sunderam, "The Evolution of the PVM Concurrent Computing System," Proceedings of 26th IEEE COMPCON Symposium, pp. 471-478, San Fransisco, February 1993.

No context found.

G.A. Geist and V.S. Sunderam, The Evolution of the PVM Concurrent Computing System , Proceeding -- 26th IEEE Compcon Symposium, pp. 471-478, San Francisco, February 1993.

No context found.

G.A. Geist and V.S. Sunderam, "The Evolution of the PVM Concurrent Computing System," Proc. 26th IEEE Compcon Symposium, 471-478, San Francisco USA (1993).

No context found.

G. A. Geist and V. S. Sunderam. The evolution of the PVM concurrent computing system. In IEEE Comput. Soc. Intl. Conf. (COMPCON), number 38, pages 549--557, Feb 1993.

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