Pub Date : 1998-06-15DOI: 10.1109/MPPOI.1998.682142
E. Harder, Hyeong-Ah Choi
We address the problem of non-preemptive scheduling of arbitrary sized file transfers in wavelength division multiplexed networks. We wish to assign a wavelength and a starting time to each path corresponding to a file in such a way that no two paths are assigned the same wavelength if concurrent transfer of files share a directed physical link. The objective is to minimize the total time spent transferring files across the network by the path set. In this paper, we obtain lower bounds on the time required for an optimal schedule and show the NP-completeness of the problem under various conditions. We then develop a scheme to produce a polynomial time approximation schedule that can be applied to arbitrary network topologies when the length of each path is bounded by two. Using this algorithm, we discuss the problem when the network topologies are rings, stars, trees, and meshes, and present near-optimal solutions.
{"title":"Scheduling file transfers in WDM optical networks","authors":"E. Harder, Hyeong-Ah Choi","doi":"10.1109/MPPOI.1998.682142","DOIUrl":"https://doi.org/10.1109/MPPOI.1998.682142","url":null,"abstract":"We address the problem of non-preemptive scheduling of arbitrary sized file transfers in wavelength division multiplexed networks. We wish to assign a wavelength and a starting time to each path corresponding to a file in such a way that no two paths are assigned the same wavelength if concurrent transfer of files share a directed physical link. The objective is to minimize the total time spent transferring files across the network by the path set. In this paper, we obtain lower bounds on the time required for an optimal schedule and show the NP-completeness of the problem under various conditions. We then develop a scheme to produce a polynomial time approximation schedule that can be applied to arbitrary network topologies when the length of each path is bounded by two. Using this algorithm, we discuss the problem when the network topologies are rings, stars, trees, and meshes, and present near-optimal solutions.","PeriodicalId":248808,"journal":{"name":"Proceedings. Fifth International Conference on Massively Parallel Processing (Cat. No.98EX182)","volume":"43 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115173264","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 : 1998-06-15DOI: 10.1109/MPPOI.1998.682121
M.M. Chang
He received both his M.S. and Ph.D. from Caltech, and a B.S. in electrical engineering with highest honors from the University of Illinois. He worked briefly as a research engineer before joining Newport Corporation, then a startup company. He was with the company for seventeen years, became its president and took the company public in 1983. He is a Fellow of the Optical Society of America, IEEE/LEOS Board of Governors, and a member of the Visiting Committee on Advanced Technology of the National Institute of Standards and Technology .
他获得了加州理工学院的硕士和博士学位，以及伊利诺伊大学(University of Illinois)的电气工程学士学位。在加入新港公司(Newport Corporation)之前，他曾短暂担任研究工程师。他在公司工作了17年，成为总裁，并于1983年将公司上市。他是美国光学学会、IEEE/LEOS理事会的成员，也是美国国家标准与技术研究所先进技术访问委员会的成员。
{"title":"Entrepeneurship is a Career Option","authors":"M.M. Chang","doi":"10.1109/MPPOI.1998.682121","DOIUrl":"https://doi.org/10.1109/MPPOI.1998.682121","url":null,"abstract":"He received both his M.S. and Ph.D. from Caltech, and a B.S. in electrical engineering with highest honors from the University of Illinois. He worked briefly as a research engineer before joining Newport Corporation, then a startup company. He was with the company for seventeen years, became its president and took the company public in 1983. He is a Fellow of the Optical Society of America, IEEE/LEOS Board of Governors, and a member of the Visiting Committee on Advanced Technology of the National Institute of Standards and Technology .","PeriodicalId":248808,"journal":{"name":"Proceedings. Fifth International Conference on Massively Parallel Processing (Cat. No.98EX182)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123398186","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 : 1998-06-15DOI: 10.1109/MPPOI.1998.682146
L. Bergman, C. Yeh, J. Morookian
At the MPPOI '96 and '97 Conferences, a new way to dynamically control in-flight pulses by a shepherd pulse to enhance time-alignment of co-propagating pulses in a bit-parallel WDM system for a single-mode fiber was discussed and the first experimental evidence that this pulse shepherding effect can be observed in a commercially available DS (dispersion-shifted) fiber was also presented. Here, we shall discuss the initial results towards the realization of a multi-km/spl times/gbytes/sec bit-parallel WDM single fiber link. The distance-speed product of this single fiber link is more than several orders of magnitude higher than that of a fiber ribbon link. The design of a 12 bit-parallel channels WDM system operating at 1 Gbit/sec per channel rate will first be presented. Experimental results for a two channel system operating at that rate will then be given. Also, new computer simulation results on how a large amplitude shepherd pulse may induce pulse compression on all the co-propagating data pulses, thereby improving the shaping of these pulses for a WDM system, will be presented and discussed.
{"title":"Towards the realization of multi-km x gbytes/sec bit-parallel WDM single fiber computer links","authors":"L. Bergman, C. Yeh, J. Morookian","doi":"10.1109/MPPOI.1998.682146","DOIUrl":"https://doi.org/10.1109/MPPOI.1998.682146","url":null,"abstract":"At the MPPOI '96 and '97 Conferences, a new way to dynamically control in-flight pulses by a shepherd pulse to enhance time-alignment of co-propagating pulses in a bit-parallel WDM system for a single-mode fiber was discussed and the first experimental evidence that this pulse shepherding effect can be observed in a commercially available DS (dispersion-shifted) fiber was also presented. Here, we shall discuss the initial results towards the realization of a multi-km/spl times/gbytes/sec bit-parallel WDM single fiber link. The distance-speed product of this single fiber link is more than several orders of magnitude higher than that of a fiber ribbon link. The design of a 12 bit-parallel channels WDM system operating at 1 Gbit/sec per channel rate will first be presented. Experimental results for a two channel system operating at that rate will then be given. Also, new computer simulation results on how a large amplitude shepherd pulse may induce pulse compression on all the co-propagating data pulses, thereby improving the shaping of these pulses for a WDM system, will be presented and discussed.","PeriodicalId":248808,"journal":{"name":"Proceedings. Fifth International Conference on Massively Parallel Processing (Cat. No.98EX182)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114337745","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 : 1998-06-15DOI: 10.1109/MPPOI.1998.682135
T.S. Jones, A. Louri
Recent advances in computer processing speeds have resulted in a parallel processing environment in which the interconnection networks (INs) themselves are the limiting factor in terms of performance. Larger and faster INs can be implemented optically subject to the current limitation in the number of wavelengths imposed by optical switch and filter technology if the INs are arranged hierarchically, i.e. if the processing nodes are arranged in clusters and these clusters are connected together special routing nodes at one or more higher levels. HORN uses optically connected rings as the basic building blocks and connects these rings using an optical tree. Three key issues facing HORN are addressed in this paper: dynamic channel allocation (DCA), optical power budget (OPE) and bit error rate (BER). Four approaches to DCA are evaluated and a design trade-off is performed between them. The first two are taken from the literature while the last two are proposed here and exploit the multiple paths available in a hierarchical network. The evaluation of OPE and BER shows that HORN is feasible and practical when optical amplification is used at the initial signal insertion point for transmissions at higher levels in the hierarchy.
{"title":"Channel allocation, power budget and bit error rate in hierarchical optical ring interconnection network (HORN)","authors":"T.S. Jones, A. Louri","doi":"10.1109/MPPOI.1998.682135","DOIUrl":"https://doi.org/10.1109/MPPOI.1998.682135","url":null,"abstract":"Recent advances in computer processing speeds have resulted in a parallel processing environment in which the interconnection networks (INs) themselves are the limiting factor in terms of performance. Larger and faster INs can be implemented optically subject to the current limitation in the number of wavelengths imposed by optical switch and filter technology if the INs are arranged hierarchically, i.e. if the processing nodes are arranged in clusters and these clusters are connected together special routing nodes at one or more higher levels. HORN uses optically connected rings as the basic building blocks and connects these rings using an optical tree. Three key issues facing HORN are addressed in this paper: dynamic channel allocation (DCA), optical power budget (OPE) and bit error rate (BER). Four approaches to DCA are evaluated and a design trade-off is performed between them. The first two are taken from the literature while the last two are proposed here and exploit the multiple paths available in a hierarchical network. The evaluation of OPE and BER shows that HORN is feasible and practical when optical amplification is used at the initial signal insertion point for transmissions at higher levels in the hierarchy.","PeriodicalId":248808,"journal":{"name":"Proceedings. Fifth International Conference on Massively Parallel Processing (Cat. No.98EX182)","volume":"242 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131570924","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 : 1998-06-15DOI: 10.1109/MPPOI.1998.682133
T. Tabata, S. Lin, S. Lee
We analyzed the performance improvement of parallel processing systems with fully or highly connected optical interconnections over systems with a bus. Solving an 1024 by 1024 point 2-D FFT with 300Mflops processors and 66MHz interconnections, the speedup factor of the execution time of fully interconnected systems over systems with a bus is 3.1 for 32 PEs, and 1.2 for 4 PEs. We proposed an extended ring system as a highly connected interconnection, and also discussed the speed-up factor as for highly connected interconnections.
{"title":"Performance improvement of parallel processing systems with fully or highly connected optical interconnections over systems with a bus","authors":"T. Tabata, S. Lin, S. Lee","doi":"10.1109/MPPOI.1998.682133","DOIUrl":"https://doi.org/10.1109/MPPOI.1998.682133","url":null,"abstract":"We analyzed the performance improvement of parallel processing systems with fully or highly connected optical interconnections over systems with a bus. Solving an 1024 by 1024 point 2-D FFT with 300Mflops processors and 66MHz interconnections, the speedup factor of the execution time of fully interconnected systems over systems with a bus is 3.1 for 32 PEs, and 1.2 for 4 PEs. We proposed an extended ring system as a highly connected interconnection, and also discussed the speed-up factor as for highly connected interconnections.","PeriodicalId":248808,"journal":{"name":"Proceedings. Fifth International Conference on Massively Parallel Processing (Cat. No.98EX182)","volume":"104 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116137657","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 : 1998-06-15DOI: 10.1109/MPPOI.1998.682143
Ge Zhou, Wencai Jing, Yimo Zhang, Feng Hua, Jindong Tian, Xiangyu Chen, Wei Liu
An optical interconnection network with parallel data transmitted through TDM optical link has been developed. The network has been applied to connect PCI bus to make up a computer cluster. A link having 20 Bits/40 Bits input and 20Bits/40Bits output in parallel and adopting a pair of fibers to transmit bi-directional data has also been implemented. An all optical network combining the TDM link and route switching by different wavelength is proposed, in which data sent to different destination computer is transmitted by optical signal at different wavelength determined by source computers. The network switch hub distinguishes wavelength and transports the lightwave to designated terminals. It is transparent to all data stream and no additional latency is inserted. This technique can be applied to establish interconnection network for the massively parallel computer and high performance computer cluster.
提出了一种通过时分复用光链路传输并行数据的光互连网络。该网络已被应用于连接PCI总线组成计算机集群。还实现了20 bit /40 bit并行输入和20 bit /40 bit并行输出，采用一对光纤进行双向数据传输的链路。提出了一种TDM链路与不同波长路由交换相结合的全光网络，将发送到不同目的计算机的数据通过源计算机确定的不同波长的光信号进行传输。网络交换机集线器识别波长并将光波传输到指定的终端。它对所有数据流是透明的，并且不会插入额外的延迟。该技术可用于建立大规模并行计算机和高性能计算机集群的互联网络。
{"title":"Design of Giga-Bit/s optical interconnection network for computer cluster","authors":"Ge Zhou, Wencai Jing, Yimo Zhang, Feng Hua, Jindong Tian, Xiangyu Chen, Wei Liu","doi":"10.1109/MPPOI.1998.682143","DOIUrl":"https://doi.org/10.1109/MPPOI.1998.682143","url":null,"abstract":"An optical interconnection network with parallel data transmitted through TDM optical link has been developed. The network has been applied to connect PCI bus to make up a computer cluster. A link having 20 Bits/40 Bits input and 20Bits/40Bits output in parallel and adopting a pair of fibers to transmit bi-directional data has also been implemented. An all optical network combining the TDM link and route switching by different wavelength is proposed, in which data sent to different destination computer is transmitted by optical signal at different wavelength determined by source computers. The network switch hub distinguishes wavelength and transports the lightwave to designated terminals. It is transparent to all data stream and no additional latency is inserted. This technique can be applied to establish interconnection network for the massively parallel computer and high performance computer cluster.","PeriodicalId":248808,"journal":{"name":"Proceedings. Fifth International Conference on Massively Parallel Processing (Cat. No.98EX182)","volume":"104 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116171012","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 : 1998-06-15DOI: 10.1109/MPPOI.1998.682134
R. Chamberlain, M. Franklin, R. Krchnavek, B.H. Baysal
This paper presents the design of an optically interconnected multiprocessor. The design is oriented to applications where the performance is bandwidth limited in conventional multiprocessors. The system utilizes board-level polymer waveguides to reduce manufacturing costs. The processor interconnection network, called Gemini, has a Banyan topology and is composed of dual optical and electronic networks. The optical data paths (circuit switched) are used for passing large data blocks and the matched electrical data paths (packet switched) are used for control of the optical interconnect and for short data messages.
{"title":"Design of an optically-interconnected multiprocessor","authors":"R. Chamberlain, M. Franklin, R. Krchnavek, B.H. Baysal","doi":"10.1109/MPPOI.1998.682134","DOIUrl":"https://doi.org/10.1109/MPPOI.1998.682134","url":null,"abstract":"This paper presents the design of an optically interconnected multiprocessor. The design is oriented to applications where the performance is bandwidth limited in conventional multiprocessors. The system utilizes board-level polymer waveguides to reduce manufacturing costs. The processor interconnection network, called Gemini, has a Banyan topology and is composed of dual optical and electronic networks. The optical data paths (circuit switched) are used for passing large data blocks and the matched electrical data paths (packet switched) are used for control of the optical interconnect and for short data messages.","PeriodicalId":248808,"journal":{"name":"Proceedings. Fifth International Conference on Massively Parallel Processing (Cat. No.98EX182)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129910785","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 : 1998-06-15DOI: 10.1109/MPPOI.1998.682120
W. Dally, Ming-Ju Edward Lee, F. An, J. Poulton, S. Tell
This paper reviews the technology of high-performance electrical signaling, presents the current state of the art, and projects future directions. We have demonstrated equalized electrical signaling between CMOS integrated circuits at data rates of 4Gb/s. As the factors that determine this signaling rate all scale with improving technology we expect the data rates of high-performance electrical signaling systems to improve on a Moore's Law curve. The frequency-dependent attenuation of copper wires sets a bandwidth-distance squared (Bd/sup 2/) limit on the distance one can signal at a given data rate. Equalizing the channel cancels inter-symbol interference caused by this attenuation and greatly increases signaling distance. In the limit of perfect equalization, distance is ultimately limited by thermal noise in the receiver. At this limit, we calculate that a 4Gb/s system will be capable of operating over 100m of 24-gauge cable without repeaters.
{"title":"High-performance electrical signaling","authors":"W. Dally, Ming-Ju Edward Lee, F. An, J. Poulton, S. Tell","doi":"10.1109/MPPOI.1998.682120","DOIUrl":"https://doi.org/10.1109/MPPOI.1998.682120","url":null,"abstract":"This paper reviews the technology of high-performance electrical signaling, presents the current state of the art, and projects future directions. We have demonstrated equalized electrical signaling between CMOS integrated circuits at data rates of 4Gb/s. As the factors that determine this signaling rate all scale with improving technology we expect the data rates of high-performance electrical signaling systems to improve on a Moore's Law curve. The frequency-dependent attenuation of copper wires sets a bandwidth-distance squared (Bd/sup 2/) limit on the distance one can signal at a given data rate. Equalizing the channel cancels inter-symbol interference caused by this attenuation and greatly increases signaling distance. In the limit of perfect equalization, distance is ultimately limited by thermal noise in the receiver. At this limit, we calculate that a 4Gb/s system will be capable of operating over 100m of 24-gauge cable without repeaters.","PeriodicalId":248808,"journal":{"name":"Proceedings. Fifth International Conference on Massively Parallel Processing (Cat. No.98EX182)","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131135915","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 : 1998-06-15DOI: 10.1109/MPPOI.1998.682145
D. Kuchta, J. Crow, P. Pepeljugoski, K. Stawiasz, J. Trewhella, D. Booth, W. Nation, C. DeCusatis, A. Muszynski
Fiber optic links offer several advantages over copper-based links including greater transmission distances, reduced cable and connector bulk and improved electrical isolation get copper-based links are still the predominant choice for parallel interconnections. The overwhelming disadvantage for fiber optic links to date has been the cost of the technology. The Jitney Optical Bus was designed to push the limits of low cost optics while maintaining enough performance for high speed parallel computing. Jitney is a 20 channel (500 Mb/s/ch) optical bus packaged in a plastic molded lead frame with a 'snap together' plastic optical coupler. The cables use plastic ferrules and large core fiber. Manufacturing cost estimates are competitive with copper and significantly under $100/Gbit/sec, for full duplex, Gigabyte/sec operation. Jitney has been successfully demonstrated in IBM AS/400 and RS6000 Power Parallel systems testbeds.
{"title":"Low cost 10 gigabit/s optical interconnects for parallel processing","authors":"D. Kuchta, J. Crow, P. Pepeljugoski, K. Stawiasz, J. Trewhella, D. Booth, W. Nation, C. DeCusatis, A. Muszynski","doi":"10.1109/MPPOI.1998.682145","DOIUrl":"https://doi.org/10.1109/MPPOI.1998.682145","url":null,"abstract":"Fiber optic links offer several advantages over copper-based links including greater transmission distances, reduced cable and connector bulk and improved electrical isolation get copper-based links are still the predominant choice for parallel interconnections. The overwhelming disadvantage for fiber optic links to date has been the cost of the technology. The Jitney Optical Bus was designed to push the limits of low cost optics while maintaining enough performance for high speed parallel computing. Jitney is a 20 channel (500 Mb/s/ch) optical bus packaged in a plastic molded lead frame with a 'snap together' plastic optical coupler. The cables use plastic ferrules and large core fiber. Manufacturing cost estimates are competitive with copper and significantly under $100/Gbit/sec, for full duplex, Gigabyte/sec operation. Jitney has been successfully demonstrated in IBM AS/400 and RS6000 Power Parallel systems testbeds.","PeriodicalId":248808,"journal":{"name":"Proceedings. Fifth International Conference on Massively Parallel Processing (Cat. No.98EX182)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133256635","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 : 1998-06-15DOI: 10.1109/MPPOI.1998.682123
E. Frietman, F. Zhao
In this paper a prototype of an optical multi-faceted Free-Space Image Distributor for optical interconnection in massively parallel processing is presented. The consecutive aspects of the design of a free space oriented image distributor are considered. Features of free space image distributing techniques are studied and solutions are proposed that favour the development, prototyping and testing of a free space oriented 9-faceted image distributing system (the Kaleidoscope) for massively parallel processing. Optically modelling the kaleidoscopic system and simulating the optical backplane are carried out. Multiple complementary alternatives in designing a multi-faceted (N>9) kaleidoscopic system are analysed.
{"title":"Optical multi-faceted free-space image distributor for massively parallel processing","authors":"E. Frietman, F. Zhao","doi":"10.1109/MPPOI.1998.682123","DOIUrl":"https://doi.org/10.1109/MPPOI.1998.682123","url":null,"abstract":"In this paper a prototype of an optical multi-faceted Free-Space Image Distributor for optical interconnection in massively parallel processing is presented. The consecutive aspects of the design of a free space oriented image distributor are considered. Features of free space image distributing techniques are studied and solutions are proposed that favour the development, prototyping and testing of a free space oriented 9-faceted image distributing system (the Kaleidoscope) for massively parallel processing. Optically modelling the kaleidoscopic system and simulating the optical backplane are carried out. Multiple complementary alternatives in designing a multi-faceted (N>9) kaleidoscopic system are analysed.","PeriodicalId":248808,"journal":{"name":"Proceedings. Fifth International Conference on Massively Parallel Processing (Cat. No.98EX182)","volume":"59 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114178099","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：