Pub Date : 1990-04-08DOI: 10.1109/DMCC.1990.555424
M. J. Gorrod, J. Kastner, Malcolm J. Coe
A m o d e l f o r t h e infra-red emis s ion f r o m t h e circumstel lar disc of a B e s tar i s presented. T h e s t ruc ture and o ther physical parame ter s of t h i s disc can be adju s t ed t o invest igate t h e infro,-red and optical lane emis s i o n f r o m such a n envelope. T h e m o d e l present ly u n d e r invest igat ion i s based o n the early work of Drnke and has been computed o n a p VAXII and a M e i k o C o m p u t i ng Surface. T h e parallel i m p l e m e n t a t i o n of t h i s mode l allows a m o r e complex and realist ic s t ruc ture t o be m o d elled in a reasonable t imescale . B o t h a n algori thmic and event decomposi t ion of t h e code have been investigated and t h e t w o m e t h o d s are compared. T h e mode l has been applied t o several B e s tar s with good agreement w i t h observat ional data.
我展示的是一张唱片。这张唱片可能会迎合我的口味。T h e m - o d e l代表ly u n d e r i s基于o离子投资igat n Drnke之早期工作已被断层扫描o n a p VAXII和a e i o k C o m p u T我ng地面。T h e平行i m p l e e n T a T i o n T h i s模式的洛杉矶allows m - o r e真ruc的情结和ic realist s T T T o d成为了m - o elled in a合理imescale。一直在调查和调查关于th e代码的事件。我一直使用T h e模式，表示良好的协调和数据。
{"title":"Transputer Modelling of Be Star Circumstellar Discs","authors":"M. J. Gorrod, J. Kastner, Malcolm J. Coe","doi":"10.1109/DMCC.1990.555424","DOIUrl":"https://doi.org/10.1109/DMCC.1990.555424","url":null,"abstract":"A m o d e l f o r t h e infra-red emis s ion f r o m t h e circumstel lar disc of a B e s tar i s presented. T h e s t ruc ture and o ther physical parame ter s of t h i s disc can be adju s t ed t o invest igate t h e infro,-red and optical lane emis s i o n f r o m such a n envelope. T h e m o d e l present ly u n d e r invest igat ion i s based o n the early work of Drnke and has been computed o n a p VAXII and a M e i k o C o m p u t i ng Surface. T h e parallel i m p l e m e n t a t i o n of t h i s mode l allows a m o r e complex and realist ic s t ruc ture t o be m o d elled in a reasonable t imescale . B o t h a n algori thmic and event decomposi t ion of t h e code have been investigated and t h e t w o m e t h o d s are compared. T h e mode l has been applied t o several B e s tar s with good agreement w i t h observat ional data.","PeriodicalId":204431,"journal":{"name":"Proceedings of the Fifth Distributed Memory Computing Conference, 1990.","volume":"24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-04-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133731048","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}
Pub Date : 1990-04-08DOI: 10.1109/DMCC.1990.555359
Izendert Huisman, Indira Nair, I. B. M. T. J. Watson, Raja Daoud
A new parallelization technique for Fault Simulation is described that is suited for message passing based parallel processors. The problem is parallelized by first casting it in Dataflow form and then constructing a Dataflow emulator for message passing systems. A fault simulator for combinational logic has been implemented on a Transputer based parallel processor, the IBM VICTOR multiprocessor. Overall performance has been measured for several logic designs.
{"title":"Fault Simulation on Message Passing","authors":"Izendert Huisman, Indira Nair, I. B. M. T. J. Watson, Raja Daoud","doi":"10.1109/DMCC.1990.555359","DOIUrl":"https://doi.org/10.1109/DMCC.1990.555359","url":null,"abstract":"A new parallelization technique for Fault Simulation is described that is suited for message passing based parallel processors. The problem is parallelized by first casting it in Dataflow form and then constructing a Dataflow emulator for message passing systems. A fault simulator for combinational logic has been implemented on a Transputer based parallel processor, the IBM VICTOR multiprocessor. Overall performance has been measured for several logic designs.","PeriodicalId":204431,"journal":{"name":"Proceedings of the Fifth Distributed Memory Computing Conference, 1990.","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-04-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130163555","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}
Pub Date : 1990-04-08DOI: 10.1109/DMCC.1990.556315
A. Chien, W. Dally
Programming languages for massively parallel concurrent computers need multi-access data abstraction tools. Most concurrent object-oriented languages serialize hierarchical abstractions. Thus multiple levels of abstraction can result in greatly diminished concurrency, even if each level only causes a tiny amount of serialization. This leaves programmers with the choice of reduced concurrency or working without useful levels of abstraction. Going without these levels of abstraction makes programs more difficult to write, understand, and debug.
{"title":"Experience with Concurrent Aggregates (CA): Implementation and Programming","authors":"A. Chien, W. Dally","doi":"10.1109/DMCC.1990.556315","DOIUrl":"https://doi.org/10.1109/DMCC.1990.556315","url":null,"abstract":"Programming languages for massively parallel concurrent computers need multi-access data abstraction tools. Most concurrent object-oriented languages serialize hierarchical abstractions. Thus multiple levels of abstraction can result in greatly diminished concurrency, even if each level only causes a tiny amount of serialization. This leaves programmers with the choice of reduced concurrency or working without useful levels of abstraction. Going without these levels of abstraction makes programs more difficult to write, understand, and debug.","PeriodicalId":204431,"journal":{"name":"Proceedings of the Fifth Distributed Memory Computing Conference, 1990.","volume":"24 3","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-04-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114033975","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}
Pub Date : 1990-04-08DOI: 10.1109/DMCC.1990.556332
T. Lewis, H. El-Rewini, J. Chu, P. Fortner, W. Su
Task Grapher is a tool for studying optimal liarallel program task scheduling on arbitrarily intercor iected parallel processors. Given a parallel prl )gram represented as a precedence-constrained task gra] h, and an interconnect topology of a target machine Task Grapher produces the following displays: 1) Gantt Chart Schedule, 2) Speed-up Line Graph, 3) ( 'ritical Path In Task Graph, 4) Processor Utilization C1 art, 5 ) Processor Efficiency Chart, 6 ) Dynamic A :tivity Display. Task Grapher currently incorporates seven scheduling heuristics which consider variabl : task execution times, variable communication lelays between processors, and target machine interconi ection topology. Task duplication heuristics allow tri de-off between space and time when running the parallel program.
{"title":"Task Grapher: A Tool for Scheduling Parallel Program Tasks","authors":"T. Lewis, H. El-Rewini, J. Chu, P. Fortner, W. Su","doi":"10.1109/DMCC.1990.556332","DOIUrl":"https://doi.org/10.1109/DMCC.1990.556332","url":null,"abstract":"Task Grapher is a tool for studying optimal liarallel program task scheduling on arbitrarily intercor iected parallel processors. Given a parallel prl )gram represented as a precedence-constrained task gra] h, and an interconnect topology of a target machine Task Grapher produces the following displays: 1) Gantt Chart Schedule, 2) Speed-up Line Graph, 3) ( 'ritical Path In Task Graph, 4) Processor Utilization C1 art, 5 ) Processor Efficiency Chart, 6 ) Dynamic A :tivity Display. Task Grapher currently incorporates seven scheduling heuristics which consider variabl : task execution times, variable communication lelays between processors, and target machine interconi ection topology. Task duplication heuristics allow tri de-off between space and time when running the parallel program.","PeriodicalId":204431,"journal":{"name":"Proceedings of the Fifth Distributed Memory Computing Conference, 1990.","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-04-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114742763","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}
Pub Date : 1990-04-08DOI: 10.1109/DMCC.1990.555381
A. Sarwal, J. Ramanathan, D. Parker, Jiang Wu
One important reason for the lack of acceptance of coronary arteriography techniques is the four hours (approximate) required for accurate 3D quantification of an arterial tree, thus making it prohibitive for the clinician to utilize. Parallel processing techniques can greatly speed up the image processing and analysis of 3D arterial trees. It has been demonstrated in this project1 that the reconstruction of the three dimensional image and arteriographic measurements can be made close to real time using these techniques. 1. Problem Description
{"title":"Parallel Processing Applied to 3D Coronary Arteriography","authors":"A. Sarwal, J. Ramanathan, D. Parker, Jiang Wu","doi":"10.1109/DMCC.1990.555381","DOIUrl":"https://doi.org/10.1109/DMCC.1990.555381","url":null,"abstract":"One important reason for the lack of acceptance of coronary arteriography techniques is the four hours (approximate) required for accurate 3D quantification of an arterial tree, thus making it prohibitive for the clinician to utilize. Parallel processing techniques can greatly speed up the image processing and analysis of 3D arterial trees. It has been demonstrated in this project1 that the reconstruction of the three dimensional image and arteriographic measurements can be made close to real time using these techniques. 1. Problem Description","PeriodicalId":204431,"journal":{"name":"Proceedings of the Fifth Distributed Memory Computing Conference, 1990.","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-04-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133835838","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}
Pub Date : 1990-04-08DOI: 10.1109/DMCC.1990.556326
P. Tseng
This paper describes the AL programming lani page for the Warp systolic array, a linear array of 11 processing cells. AL has been successfully used tc program matrix computation applications on WarIi. Examples of LU decomposition, QR decompositioii , and singular value decomposition (SVD) were giv:n in this paper to illustrate the use of AL. These AI procedures were rewritten from the LINPACK Fortran programs; their algorithms and numerical prop xties were kept the same as the LINPACK implementations. More than 27 MFLOPS (out of 100 MFI ,OPS peak) on matrices of order 300 were achieved for these procedures.
{"title":"A Systolic Array Programming Language","authors":"P. Tseng","doi":"10.1109/DMCC.1990.556326","DOIUrl":"https://doi.org/10.1109/DMCC.1990.556326","url":null,"abstract":"This paper describes the AL programming lani page for the Warp systolic array, a linear array of 11 processing cells. AL has been successfully used tc program matrix computation applications on WarIi. Examples of LU decomposition, QR decompositioii , and singular value decomposition (SVD) were giv:n in this paper to illustrate the use of AL. These AI procedures were rewritten from the LINPACK Fortran programs; their algorithms and numerical prop xties were kept the same as the LINPACK implementations. More than 27 MFLOPS (out of 100 MFI ,OPS peak) on matrices of order 300 were achieved for these procedures.","PeriodicalId":204431,"journal":{"name":"Proceedings of the Fifth Distributed Memory Computing Conference, 1990.","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-04-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132804384","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}
Pub Date : 1990-04-08DOI: 10.1109/DMCC.1990.555361
T. D. Gottschalk
is minimized over all permutations II. For NA 5 N B , the naive (exhaustive search) complexity of the assignment problem is O[NB! / (NB N A ) ! ] . There are, however, a variety of exact solutions to the assignment problem with reduced complexity O[N;NB], (Refs.[l-31). Section 2 briefly describes one such method, Munkres Algorithm [2] , and presents a particular sequential implementation. Performance of the algorithm is examined for the particularly nasty problem of associating lists of random points within the unit square. In Section 3, the algorithm is generalized for concurrent execution, and performance results for runs on the Mark111 hypercube are presented. The input to the assignment problem is the matrix D Z E { d i j } of dissimilarities from Eq.(3). The first point to note is that the particular assignment which minimizes Eq.(6) is not altered if afixed value is added to or subtracted from all entries in any row or column of the cost matrix D. Exploiting this fact, Munkres solution to the Assignment Problem can be divided into two parts
在所有排列中最小。对于n5nb，赋值问题的朴素(穷举搜索)复杂度为0 [NB!/ (nb n a) !］ . 然而，有多种精确解的分配问题，降低了复杂度O[N;NB]，(参考文献[l-31])。第2节简要介绍了一种这样的方法，Munkres算法[2]，并给出了一个特定的顺序实现。对该算法的性能进行了检验，以解决单位平方内随机点的关联列表这一特别棘手的问题。在第3节中，将该算法推广到并发执行，并给出了在Mark111超立方体上运行的性能结果。分配问题的输入是与Eq.(3)不相似的矩阵D Z E {D i j}。首先要注意的是，如果在成本矩阵d的任何行或列的所有条目中添加或减去固定值，则最小化Eq.(6)的特定分配不会改变。利用这一事实，Munkres对分配问题的解决方案可以分为两个部分
{"title":"Concurrent Implementation of Munkres Algorithm","authors":"T. D. Gottschalk","doi":"10.1109/DMCC.1990.555361","DOIUrl":"https://doi.org/10.1109/DMCC.1990.555361","url":null,"abstract":"is minimized over all permutations II. For NA 5 N B , the naive (exhaustive search) complexity of the assignment problem is O[NB! / (NB N A ) ! ] . There are, however, a variety of exact solutions to the assignment problem with reduced complexity O[N;NB], (Refs.[l-31). Section 2 briefly describes one such method, Munkres Algorithm [2] , and presents a particular sequential implementation. Performance of the algorithm is examined for the particularly nasty problem of associating lists of random points within the unit square. In Section 3, the algorithm is generalized for concurrent execution, and performance results for runs on the Mark111 hypercube are presented. The input to the assignment problem is the matrix D Z E { d i j } of dissimilarities from Eq.(3). The first point to note is that the particular assignment which minimizes Eq.(6) is not altered if afixed value is added to or subtracted from all entries in any row or column of the cost matrix D. Exploiting this fact, Munkres solution to the Assignment Problem can be divided into two parts","PeriodicalId":204431,"journal":{"name":"Proceedings of the Fifth Distributed Memory Computing Conference, 1990.","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-04-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133015454","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}
Pub Date : 1990-04-08DOI: 10.1109/DMCC.1990.555393
C. Still
This paper describes work done on the 1024 node NCUBE hypercube at the University of South Carolina in developing methods for efficient local solution of unconstrained minimization problems. The paper begins with a mathematical discussion of quasiNewton methods for unconstrained optimization, and specifically Broyden’s Method. Next it presents the paralfel methods, and discusses the parallel implementation of the most common Broyden method. Finally it lists some numerical results to evaluate the performance of the parallel Broyden methods.
{"title":"Parallel Quasi-Newton Methods for Unconstrained Optimization","authors":"C. Still","doi":"10.1109/DMCC.1990.555393","DOIUrl":"https://doi.org/10.1109/DMCC.1990.555393","url":null,"abstract":"This paper describes work done on the 1024 node NCUBE hypercube at the University of South Carolina in developing methods for efficient local solution of unconstrained minimization problems. The paper begins with a mathematical discussion of quasiNewton methods for unconstrained optimization, and specifically Broyden’s Method. Next it presents the paralfel methods, and discusses the parallel implementation of the most common Broyden method. Finally it lists some numerical results to evaluate the performance of the parallel Broyden methods.","PeriodicalId":204431,"journal":{"name":"Proceedings of the Fifth Distributed Memory Computing Conference, 1990.","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-04-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121887379","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}
Pub Date : 1990-04-08DOI: 10.1109/DMCC.1990.556287
K. Zemoudeh, A. Sengupta
{"title":"Routing Frequently Used Bijections on Hypercube","authors":"K. Zemoudeh, A. Sengupta","doi":"10.1109/DMCC.1990.556287","DOIUrl":"https://doi.org/10.1109/DMCC.1990.556287","url":null,"abstract":"","PeriodicalId":204431,"journal":{"name":"Proceedings of the Fifth Distributed Memory Computing Conference, 1990.","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-04-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126086707","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}
Pub Date : 1990-04-08DOI: 10.1109/DMCC.1990.556407
M. Peercy, P. Banerjee
In this paper we present an evaluation of communication in the network of a faulty hypercube using an approximate analytical model and a detailed simulation model. We asume the hypercube has a deterministic routing strategy. Consequently, we are able to use simple models to derive the traffic on each link. Additionally, we develop a theory with which we find the message delay not only in packet-switched hypercube networks, but in the more current circuit-switched networks as well. Thus we can examine the degradation from ideal communication performance due to different fault patterns and different reroute strategies we may use to compensate for faults. We next describe a program, EASY, which implements our approximate analysis along with simulation for detailed evaluation of message communication. We include a comparison of our theoretical evaluation with simulation. We also include a brief comparison of reroute techniques for faulty hypercubes.
{"title":"A Method for Evaluating Message Communication in Faulty Hypercubes","authors":"M. Peercy, P. Banerjee","doi":"10.1109/DMCC.1990.556407","DOIUrl":"https://doi.org/10.1109/DMCC.1990.556407","url":null,"abstract":"In this paper we present an evaluation of communication in the network of a faulty hypercube using an approximate analytical model and a detailed simulation model. We asume the hypercube has a deterministic routing strategy. Consequently, we are able to use simple models to derive the traffic on each link. Additionally, we develop a theory with which we find the message delay not only in packet-switched hypercube networks, but in the more current circuit-switched networks as well. Thus we can examine the degradation from ideal communication performance due to different fault patterns and different reroute strategies we may use to compensate for faults. We next describe a program, EASY, which implements our approximate analysis along with simulation for detailed evaluation of message communication. We include a comparison of our theoretical evaluation with simulation. We also include a brief comparison of reroute techniques for faulty hypercubes.","PeriodicalId":204431,"journal":{"name":"Proceedings of the Fifth Distributed Memory Computing Conference, 1990.","volume":"45 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1990-04-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127188593","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：