In this paper we discuss intelligent agent support for parallel and distributed computing in a heterogeneous environment. We provide an overview of the Bond environment and of services provided by a network on intelligent agents, then we discuss in depth the Scheduling Expert Advisor SEA. The SEA processes a high level description of a computational task provided by a user and converts it into a set of facts and rules. An expert system starts the execution of one program or a group of programs based on the scheduling information compiled earlier.
{"title":"A scheduling expert advisor for heterogeneous environments","authors":"Mihai G. Sirbu, D. Marinescu","doi":"10.1109/HCW.1997.581411","DOIUrl":"https://doi.org/10.1109/HCW.1997.581411","url":null,"abstract":"In this paper we discuss intelligent agent support for parallel and distributed computing in a heterogeneous environment. We provide an overview of the Bond environment and of services provided by a network on intelligent agents, then we discuss in depth the Scheduling Expert Advisor SEA. The SEA processes a high level description of a computational task provided by a user and converts it into a set of facts and rules. An expert system starts the execution of one program or a group of programs based on the scheduling information compiled earlier.","PeriodicalId":286909,"journal":{"name":"Proceedings Sixth Heterogeneous Computing Workshop (HCW'97)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122827561","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
P. Messina, S. Brunett, D. Davis, T. D. Gottschalk, D. Curkendall, L. Ekroot, H. Siegel
Interactive simulation of battles is a valuable tool for training. The behavior and movement of hundreds or thousands of entities (tanks, trucks, airplanes, missiles, etc.) is currently simulated using dozens or more workstations on geographically distributed LANs connected by WANs. The simulated entities can move, fire weapons, receive "radio" messages, etc. The terrain that they traverse may change dynamically, for example due to rains turning dirt roads into mud or bombs forming craters. Thus the entities need to receive frequent information about the state of the terrain and the location and state of other entities. Typically, information is updated several times a second. As the number of simulated entities grows, the number of messages that need to be sent per unit of time can grow to unmanageable numbers. One approach to reducing the number of messages is to keep track of what entities need to know about which other entities and only send information to the entities that need to know. For example, tanks in Germany need not know about a change of course of a ship in the Pacific. This technique for reducing messages is known as interest management. Caltech and its Jet Propulsion Laboratory have implemented a simulation of this type on several large-scale parallel computers, exploiting both the compute power and the fast messaging fabric of such systems. The application is implemented using a heterogeneous approach. Some nodes are used to simulate entities, some to manage a database of terrain information, some to provide interest management functions, and some to route messages to the entities that do need to receive the information. Some of these tasks require more memory than others, some require faster processing capability.
{"title":"Distributed interactive simulation for synthetic forces","authors":"P. Messina, S. Brunett, D. Davis, T. D. Gottschalk, D. Curkendall, L. Ekroot, H. Siegel","doi":"10.1109/HCW.1997.581414","DOIUrl":"https://doi.org/10.1109/HCW.1997.581414","url":null,"abstract":"Interactive simulation of battles is a valuable tool for training. The behavior and movement of hundreds or thousands of entities (tanks, trucks, airplanes, missiles, etc.) is currently simulated using dozens or more workstations on geographically distributed LANs connected by WANs. The simulated entities can move, fire weapons, receive \"radio\" messages, etc. The terrain that they traverse may change dynamically, for example due to rains turning dirt roads into mud or bombs forming craters. Thus the entities need to receive frequent information about the state of the terrain and the location and state of other entities. Typically, information is updated several times a second. As the number of simulated entities grows, the number of messages that need to be sent per unit of time can grow to unmanageable numbers. One approach to reducing the number of messages is to keep track of what entities need to know about which other entities and only send information to the entities that need to know. For example, tanks in Germany need not know about a change of course of a ship in the Pacific. This technique for reducing messages is known as interest management. Caltech and its Jet Propulsion Laboratory have implemented a simulation of this type on several large-scale parallel computers, exploiting both the compute power and the fast messaging fabric of such systems. The application is implemented using a heterogeneous approach. Some nodes are used to simulate entities, some to manage a database of terrain information, some to provide interest management functions, and some to route messages to the entities that do need to receive the information. Some of these tasks require more memory than others, some require faster processing capability.","PeriodicalId":286909,"journal":{"name":"Proceedings Sixth Heterogeneous Computing Workshop (HCW'97)","volume":"60 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115894269","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The problem of statically estimating the execution time distribution for a task graph consisting of a collection of subtasks to be executed in a heterogeneous computing (HC) system is considered. Execution time distributions for the individual subtasks are assumed to be known. A mathematical model for the communication network: that interconnects the machines of the HC system is introduced and a probabilistic approach is developed to estimate the overall execution time distribution of the task graph. It is shown that, for a given matching and scheduling, computing the exact distribution of the overall execution time of a task graph is very difficult, and thus impractical. The proposed approach approximates the exact distribution and requires a relatively small amount of calculation time. The accuracy of the proposed approach is demonstrated mathematically through the derivation of bounds that quantify the difference between the exact distribution and that provided by the proposed approach. Numerical studies are also included to further validate the utility of the proposed methodology.
{"title":"Estimating the execution time distribution for a task graph in a heterogeneous computing system","authors":"Y. Li, J. Antonio","doi":"10.1109/HCW.1997.581419","DOIUrl":"https://doi.org/10.1109/HCW.1997.581419","url":null,"abstract":"The problem of statically estimating the execution time distribution for a task graph consisting of a collection of subtasks to be executed in a heterogeneous computing (HC) system is considered. Execution time distributions for the individual subtasks are assumed to be known. A mathematical model for the communication network: that interconnects the machines of the HC system is introduced and a probabilistic approach is developed to estimate the overall execution time distribution of the task graph. It is shown that, for a given matching and scheduling, computing the exact distribution of the overall execution time of a task graph is very difficult, and thus impractical. The proposed approach approximates the exact distribution and requires a relatively small amount of calculation time. The accuracy of the proposed approach is demonstrated mathematically through the derivation of bounds that quantify the difference between the exact distribution and that provided by the proposed approach. Numerical studies are also included to further validate the utility of the proposed methodology.","PeriodicalId":286909,"journal":{"name":"Proceedings Sixth Heterogeneous Computing Workshop (HCW'97)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117053323","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In a dedicated mixed-machine heterogeneous computing (HC) system, an application program may be decomposed into subtasks, then each subtask assigned to the machine where it is best suited for execution. Subtask data relocation is defined as selecting the sources for their needed data items. This study focuses on theoretical issues for data relocation using a stochastic HC model. It is assumed that multiple independent subtasks of an application program can be executed concurrently on different machines whenever possible. A stochastic model for HC is proposed, in which the computation times of subtasks and communication times for inter-machine data transfers can be random variables. The optimization problem for finding the optimal matching, scheduling, and data relocation schemes to minimize the total execution time of an application program is defined based on this stochastic HC model. The optimization criteria and search space for the above optimization problem are described. It is proven that a greedy algorithm based approach will generate the optimal data relocation scheme with respect to any fixed matching and scheduling schemes. This result indicates that a greedy algorithm based approach is the best strategy for developing data relocation heuristics in practice.
{"title":"A stochastic model of a dedicated heterogeneous computing system for establishing a greedy approach to developing data relocation heuristics","authors":"Min Tan, H. Siegel","doi":"10.1109/HCW.1997.581415","DOIUrl":"https://doi.org/10.1109/HCW.1997.581415","url":null,"abstract":"In a dedicated mixed-machine heterogeneous computing (HC) system, an application program may be decomposed into subtasks, then each subtask assigned to the machine where it is best suited for execution. Subtask data relocation is defined as selecting the sources for their needed data items. This study focuses on theoretical issues for data relocation using a stochastic HC model. It is assumed that multiple independent subtasks of an application program can be executed concurrently on different machines whenever possible. A stochastic model for HC is proposed, in which the computation times of subtasks and communication times for inter-machine data transfers can be random variables. The optimization problem for finding the optimal matching, scheduling, and data relocation schemes to minimize the total execution time of an application program is defined based on this stochastic HC model. The optimization criteria and search space for the above optimization problem are described. It is proven that a greedy algorithm based approach will generate the optimal data relocation scheme with respect to any fixed matching and scheduling schemes. This result indicates that a greedy algorithm based approach is the best strategy for developing data relocation heuristics in practice.","PeriodicalId":286909,"journal":{"name":"Proceedings Sixth Heterogeneous Computing Workshop (HCW'97)","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127281191","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
UbiWorld is a concept being developed by the Futures Laboratory group at Argonne National Laboratory that ties together the notion of ubiquitous computing (Ubicomp) with that of using virtual reality for rapid prototyping. The goal is to develop an environment where one can explore Ubicomp-type concepts without having to build real Ubicomp hardware. The basic notion is to extend object models in a virtual world by using distributed wide area heterogeneous computing technology to provide complex networking and processing capabilities to virtual reality objects.
UbiWorld是由美国阿贡国家实验室(Argonne National Laboratory)的未来实验室(Futures Laboratory)开发的一个概念，它将泛在计算(Ubicomp)的概念与使用虚拟现实进行快速原型设计的概念联系在一起。我们的目标是开发一个环境，在这个环境中，人们可以探索Ubicomp类型的概念，而不必构建真正的Ubicomp硬件。其基本概念是通过使用分布式广域异构计算技术扩展虚拟世界中的对象模型，为虚拟现实对象提供复杂的网络和处理能力。
{"title":"UbiWorld: an environment integrating virtual reality, supercomputing, and design","authors":"T. Disz, M. Papka, R. Stevens","doi":"10.1109/HCW.1997.581409","DOIUrl":"https://doi.org/10.1109/HCW.1997.581409","url":null,"abstract":"UbiWorld is a concept being developed by the Futures Laboratory group at Argonne National Laboratory that ties together the notion of ubiquitous computing (Ubicomp) with that of using virtual reality for rapid prototyping. The goal is to develop an environment where one can explore Ubicomp-type concepts without having to build real Ubicomp hardware. The basic notion is to extend object models in a virtual world by using distributed wide area heterogeneous computing technology to provide complex networking and processing capabilities to virtual reality objects.","PeriodicalId":286909,"journal":{"name":"Proceedings Sixth Heterogeneous Computing Workshop (HCW'97)","volume":"66 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126557123","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
A heterogeneous multicomputer is a multicomputer composed of two or more different types of processors. This paper describes the rationale for heterogeneity in a multicomputer and gives a typical example of a heterogeneous system in the form of a RACE multicomputer composed of a mixture of Analog Devices' SHARC 21060 and the IBM/Motorola/Apple PowerPC 603p processors. These two processors have complementary attributes, and the advantages and limitations of each are described. Multicomputers generally implement a sequence of different processing algorithms. The "optimal" processor that maximizes throughput at each step in the processing flow is generally a function of the algorithm to be executed at that step. Other factors that also influence the optimal mix of processors in a heterogeneous multicomputer include physical processing density, hardware cost, and ease of programmability.
{"title":"Mercury Computer Systems' modular heterogeneous RACE(R) multicomputer","authors":"T. H. Einstein","doi":"10.1109/HCW.1997.581410","DOIUrl":"https://doi.org/10.1109/HCW.1997.581410","url":null,"abstract":"A heterogeneous multicomputer is a multicomputer composed of two or more different types of processors. This paper describes the rationale for heterogeneity in a multicomputer and gives a typical example of a heterogeneous system in the form of a RACE multicomputer composed of a mixture of Analog Devices' SHARC 21060 and the IBM/Motorola/Apple PowerPC 603p processors. These two processors have complementary attributes, and the advantages and limitations of each are described. Multicomputers generally implement a sequence of different processing algorithms. The \"optimal\" processor that maximizes throughput at each step in the processing flow is generally a function of the algorithm to be executed at that step. Other factors that also influence the optimal mix of processors in a heterogeneous multicomputer include physical processing density, hardware cost, and ease of programmability.","PeriodicalId":286909,"journal":{"name":"Proceedings Sixth Heterogeneous Computing Workshop (HCW'97)","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116691408","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We describe a novel framework for early detection and isolation of security violations in heterogeneous environments, based on real-time service hopping. In distributed client-server systems, service hopping fills a role analogous to frequency-hopping spread-spectrum techniques for secure wireless communication. The framework incorporates design principles for secure hopping, as well as engineering principles for improving throughput in the presence of a statistically noisy interconnection network. We describe potential large-scale applications of the hopping techniques, and present some initial experimental results with a hopping client-server system.
{"title":"The hopping ruse","authors":"Marina C. Chen, J. Cowie","doi":"10.1109/HCW.1997.581422","DOIUrl":"https://doi.org/10.1109/HCW.1997.581422","url":null,"abstract":"We describe a novel framework for early detection and isolation of security violations in heterogeneous environments, based on real-time service hopping. In distributed client-server systems, service hopping fills a role analogous to frequency-hopping spread-spectrum techniques for secure wireless communication. The framework incorporates design principles for secure hopping, as well as engineering principles for improving throughput in the presence of a statistically noisy interconnection network. We describe potential large-scale applications of the hopping techniques, and present some initial experimental results with a hopping client-server system.","PeriodicalId":286909,"journal":{"name":"Proceedings Sixth Heterogeneous Computing Workshop (HCW'97)","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117043612","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
D. Arapov, A. Kalinov, Alexey L. Lastovetsky, Ilya Ledovskih, T. Lewis
mpC is a medium level programming language for distributed memory machines (DMM). The language is an ANSI C superset based on the notion of a network comprising virtual processors of different types and performances connected with links of different bandwidths. It allows the user to describe a network topology, create and discard networks, and distribute data and computations over the networks. In other words, the user can specify (dynamically) the topology of his application, and the mpC programming environment will use this (topological) information in run time to ensure the efficient execution of the application on any particular DMM. The paper outlines the principal features of mpC and its programming environment which make them suitable tools to write efficient and portable parallel programs for heterogenous DMM.
{"title":"A programming environment for heterogenous distributed memory machines","authors":"D. Arapov, A. Kalinov, Alexey L. Lastovetsky, Ilya Ledovskih, T. Lewis","doi":"10.1109/HCW.1997.581408","DOIUrl":"https://doi.org/10.1109/HCW.1997.581408","url":null,"abstract":"mpC is a medium level programming language for distributed memory machines (DMM). The language is an ANSI C superset based on the notion of a network comprising virtual processors of different types and performances connected with links of different bandwidths. It allows the user to describe a network topology, create and discard networks, and distribute data and computations over the networks. In other words, the user can specify (dynamically) the topology of his application, and the mpC programming environment will use this (topological) information in run time to ensure the efficient execution of the application on any particular DMM. The paper outlines the principal features of mpC and its programming environment which make them suitable tools to write efficient and portable parallel programs for heterogenous DMM.","PeriodicalId":286909,"journal":{"name":"Proceedings Sixth Heterogeneous Computing Workshop (HCW'97)","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121490038","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
A. Reinefeld, R. Baraglia, T. Decker, J. Gehring, D. Laforenza, F. Ramme, Thomas Römke, J. Simon
Distributed high-performance computing (so-called metacomputing) refers to the coordinated use of a pool of geographically distributed high-performance computers. The user's view of an ideal metacomputer is that of a powerful monolithic virtual machine. The implementor's view, on the other hand, is that of a variety of interacting services implemented in a scalable and extensible manner. We present MOL, the Metacomputer Online environment. In contrast to other metacomputing environments, MOL is not based on specific programming models or tools. It has rather been designed as an open, extensible software system comprising a variety of software modules, each of them specialized in serving one specific task such as resource scheduling, job control, task communication, task migration, user interface, and much more. All of these modules exist and are working. The main challenge in the design of MOL lies in the specification of suitable, generic interfaces for the effective interaction between the modules.
{"title":"The MOL project: an open, extensible metacomputer","authors":"A. Reinefeld, R. Baraglia, T. Decker, J. Gehring, D. Laforenza, F. Ramme, Thomas Römke, J. Simon","doi":"10.1109/HCW.1997.581407","DOIUrl":"https://doi.org/10.1109/HCW.1997.581407","url":null,"abstract":"Distributed high-performance computing (so-called metacomputing) refers to the coordinated use of a pool of geographically distributed high-performance computers. The user's view of an ideal metacomputer is that of a powerful monolithic virtual machine. The implementor's view, on the other hand, is that of a variety of interacting services implemented in a scalable and extensible manner. We present MOL, the Metacomputer Online environment. In contrast to other metacomputing environments, MOL is not based on specific programming models or tools. It has rather been designed as an open, extensible software system comprising a variety of software modules, each of them specialized in serving one specific task such as resource scheduling, job control, task communication, task migration, user interface, and much more. All of these modules exist and are working. The main challenge in the design of MOL lies in the specification of suitable, generic interfaces for the effective interaction between the modules.","PeriodicalId":286909,"journal":{"name":"Proceedings Sixth Heterogeneous Computing Workshop (HCW'97)","volume":"28 2","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120895397","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this paper, a generic technique for mapping heterogeneous task graphs onto heterogeneous system graphs is presented. The task and system graphs studied in this paper have nonuniform computation and communication weights associated with the nodes and the edges. Two clustering algorithms have been proposed which can be used to obtain a multilayer clustered graph called a Spec graph from a given task graph and a multilayer clustered graph called a Rep graph from a given system graph. We present a mapping algorithm which produces a suboptimal matching of a given Spec graph containing M task modules, onto a Rep graph of N processors, in O(MP) fame, where P=max(M,N). Our experimental results indicate that our mapping algorithm is the fastest one and generates results which are better than, or similar to, those of other leading techniques which work only for restricted task or system graphs.
{"title":"Mapping heterogeneous task graphs onto heterogeneous system graphs","authors":"M. Eshaghian-Wilner, Ying-Chieh Wu","doi":"10.1109/HCW.1997.581417","DOIUrl":"https://doi.org/10.1109/HCW.1997.581417","url":null,"abstract":"In this paper, a generic technique for mapping heterogeneous task graphs onto heterogeneous system graphs is presented. The task and system graphs studied in this paper have nonuniform computation and communication weights associated with the nodes and the edges. Two clustering algorithms have been proposed which can be used to obtain a multilayer clustered graph called a Spec graph from a given task graph and a multilayer clustered graph called a Rep graph from a given system graph. We present a mapping algorithm which produces a suboptimal matching of a given Spec graph containing M task modules, onto a Rep graph of N processors, in O(MP) fame, where P=max(M,N). Our experimental results indicate that our mapping algorithm is the fastest one and generates results which are better than, or similar to, those of other leading techniques which work only for restricted task or system graphs.","PeriodicalId":286909,"journal":{"name":"Proceedings Sixth Heterogeneous Computing Workshop (HCW'97)","volume":"31 1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122577979","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：