{"title":"Applying optical interconnects to electronic systems. Promise vs. practicality","authors":"J. Bristow, J. Lehman, M. Hibbs-Brenner, Yue Liu","doi":"10.1109/MPPOI.1997.609094","DOIUrl":"https://doi.org/10.1109/MPPOI.1997.609094","url":null,"abstract":"Optical interconnects have long promised to solve a number of problems assumed to exist in electronic systems. A wealth of technology exists in the area, yet few electronic systems today employ optical interconnects. In many cases, it is the cost of optical interconnects which prevents their widespread adoption, as little fundamental advantage is offered by most applications of optical interconnects. The decreasing cost of optoelectronic technology and of its incorporation into electronic systems will enable further penetration of the market. Plastic fiber is likely to provide a competitive solution for intercabinet interconnects, while both high density fiber backplanes and polymer backplanes are likely to impact first high performance computer systems and eventually a broader range of systems, while free space interconnects will facilitate both advanced versions of existing architectures and the implementation of new systems. The key to practical implementation lies in providing transparency to the system integrator to minimize the departure from established practices. Significant progress is also being made in a new direction in which the inherent advantages of optics can be exploited in unique architectures, while still implementing logic functions in electronics, however such systems are unlikely to find widespread use in the near future.","PeriodicalId":406486,"journal":{"name":"Proceedings of the Fourth International Conference on Massively Parallel Processing Using Optical Interconnections","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134464171","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}
{"title":"Embedding binary trees in arrays with optical buses","authors":"Yousong Mei, C. Qiao","doi":"10.1109/MPPOI.1997.609161","DOIUrl":"https://doi.org/10.1109/MPPOI.1997.609161","url":null,"abstract":"Optical bus based architectures have received much attention lately, and how to take advantage of the unique communication capabilities of such architectures in application development needs to be demonstrated. In this paper we first describe an embedding of a binary tree into a Reconfigurable Array with Spanning Optical Buses (or RASOB), which achieves a near minimal expansion cost. We then describe an embedding of a large binary tree into a small RASOB, which achieves a near optimal load balance. In both embeddings, all packet transfers between parents and children in the binary trees can be accomplished in one bus cycle, which implies that the embeddings achieve an optimum dilation cost of 1.","PeriodicalId":406486,"journal":{"name":"Proceedings of the Fourth International Conference on Massively Parallel Processing Using Optical Interconnections","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114530395","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}
{"title":"A pipelined TDM optical bus with conditional delays","authors":"Yueming Li, Y. Pan, S. Zheng","doi":"10.1109/MPPOI.1997.609208","DOIUrl":"https://doi.org/10.1109/MPPOI.1997.609208","url":null,"abstract":"A new pipelined time-division multiplexing optical bus for implementing a linear array parallel computer architecture is proposed. In this bus system, switches are introduced on the receiving segment of the bus to control the signal delays on the optical waveguide. The states of switches are dynamically programmable under the control of processors according to computation needs. The linear processor arrays based on such buses can be used as building blocks to construct parallel architectures of higher dimensions to achieve improved scalability and performance.","PeriodicalId":406486,"journal":{"name":"Proceedings of the Fourth International Conference on Massively Parallel Processing Using Optical Interconnections","volume":"60 7","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120921803","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}
{"title":"Parallel optical interconnection for massively parallel processor RWC-1","authors":"T. Yoshikawa, H. Matsuoka, T. Yokota, J. Shimada","doi":"10.1109/MPPOI.1997.609047","DOIUrl":"https://doi.org/10.1109/MPPOI.1997.609047","url":null,"abstract":"A GByte/sec-class optical interconnection subsystem for a parallel computer was developed, and it operated stably in the testbed system of a parallel computer, Real World Computer-1 (RWC-1). Although it consisted of MMF parallel links, data was transmitted over 1 km because of deskew operation of +/- 15 ns. Random packets were transmitted without error over 17.5 hours, corresponding to a BER of 10/sup -14/.","PeriodicalId":406486,"journal":{"name":"Proceedings of the Fourth International Conference on Massively Parallel Processing Using Optical Interconnections","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122193107","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}
{"title":"Multicasting control and communications on multihop stack-ring OPS networks","authors":"Afonso Ferreira, Eric Fleury, M. Grammatikakis","doi":"10.1109/MPPOI.1997.609085","DOIUrl":"https://doi.org/10.1109/MPPOI.1997.609085","url":null,"abstract":"We propose dynamic multicasting on an N=m.n processor; OPS-based stack-ring topology consisting of n processor groups, with m processors each. By considering multicasting, we realize the increased network bandwidth offered by one-to-many optical communications. We assume that each processor has infinite local and network buffers and that the system performs synchronously. Assuming shortest path routing and network packet priority we compute network throughput and latency measures as a function of the packet arrival rate. Balanced networks with m/spl ap/n offer higher throughput for uniform, multicast-free traffic. When the multicasting load and especially the locality of references increase, stack-ring systems with n>m offer higher throughput and reasonable latency. A high level of locality and multicasting are necessary for achieving stack-ring network scalability.","PeriodicalId":406486,"journal":{"name":"Proceedings of the Fourth International Conference on Massively Parallel Processing Using Optical Interconnections","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123761353","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}
{"title":"Four multiplexed, 8/spl times/8-bit 2-D parallel transmission based upon space-CDMA","authors":"M. Nakamura, K. Kitayama, Y. Igasaki, K. Kaneda","doi":"10.1109/MPPOI.1997.609212","DOIUrl":"https://doi.org/10.1109/MPPOI.1997.609212","url":null,"abstract":"4 channel multiplexing for 64 bit two dimensional (2-D) parallel optical data link based upon optical space-CDMA is experimentally demonstrated by using newly developed modules of optical spatial encoder and decoder. Each 8/spl times/8 bit-plane (64 bit parallel data) is optically encoded with an 8/spl times/8 2-D optical orthogonal signature pattern. The encoded bit-planes are spatially multiplexed, and it is transmitted through an image fiber having 10/sup 5/ cores. Each receiver can recover the intended input bit-plane via the optical decoding process. Total error-rate is 2.73%. This result will encourage the application of optical space-CDMA to future high-throughput 2-D parallel data links connecting massively parallel processors.","PeriodicalId":406486,"journal":{"name":"Proceedings of the Fourth International Conference on Massively Parallel Processing Using Optical Interconnections","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126657139","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}
{"title":"Wavelength recognizing switches: Architectures and system applications","authors":"B. Hoanca, A. Sawchuk","doi":"10.1109/MPPOI.1997.609134","DOIUrl":"https://doi.org/10.1109/MPPOI.1997.609134","url":null,"abstract":"We present and model possible system applications of a recently demonstrated wavelength recognizing switch (WRS). The switch allows all-optical self routing in very wide bandwidth networks. In this work, we discuss architectures using multiwavelength all-optical WRS devices, we analyze possible network access strategies and we assess the feasibility of cascading WRS devices. We conclude with a novel view of a higher level of data aggregation, where the basic switching unit is wavelength packet (multiple modulated wavelengths switched as a single entity).","PeriodicalId":406486,"journal":{"name":"Proceedings of the Fourth International Conference on Massively Parallel Processing Using Optical Interconnections","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122536834","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}
{"title":"Self-routing in 2-D shuffle networks with dimension-dependent switches and interconnections","authors":"J. Giglmayr","doi":"10.1109/MPPOI.1997.609166","DOIUrl":"https://doi.org/10.1109/MPPOI.1997.609166","url":null,"abstract":"In shuffle multistage interconnection networks (MINs), by increasing the size of the switches, the number of stages (NS) and in turn the attenuation of signals decreases. An interesting concept for self-routing in shuffle MINs are dimension-dependent switches, i.e. switches of size 2/sup d//spl times/2/sup d/ (where d/spl ges/1 is the dimension of the interconnected data set) and dimension-dependent interconnections. Throughout the paper the 1-D shuffle interconnections and k/spl times/k-switches (k/spl ges/4) are replaced by d-dimensional shuffles (d/spl ges/2) and switches of size 2/sup d//spl times/2/sup d/ though the networks are still planar. In this way, given a data set, the original contribution of the paper is the presentation of (1) the number of topologically equivalent shuffle networks, (2) the complete concept of self-routing in (artificial) higher-dimensional shuffle networks (3) the solution of self-routing in 2-D shuffle networks not solvable by the common concept and (4) the decomposition of the input/output array into several small arrays without loss of functionality. The latter is important as presently reliable laser arrays have size of at most 8/spl times/8 and 16/spl times/16, respectively.","PeriodicalId":406486,"journal":{"name":"Proceedings of the Fourth International Conference on Massively Parallel Processing Using Optical Interconnections","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128286616","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}
{"title":"BPC permutations on the OTIS-Mesh optoelectronic computer","authors":"S. Sahni, Chih-Fang Wang","doi":"10.1109/MPPOI.1997.609157","DOIUrl":"https://doi.org/10.1109/MPPOI.1997.609157","url":null,"abstract":"We show that the diameter of an N/sup 2/ processor OTIS-Mesh is 4/spl radic/(N-3). Two possible embeddings of an N/spl times/N mesh onto an OTIS-Mesh are evaluated. OTIS-Mesh algorithms for some commonly performed permutations-transpose, bit reversal, vector reversal, perfect shuffle, unshuffle, shuffled row-major, and bit shuffle-are developed. We also propose an algorithm for general BPC permutations.","PeriodicalId":406486,"journal":{"name":"Proceedings of the Fourth International Conference on Massively Parallel Processing Using Optical Interconnections","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124498855","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}
S. Chai, A. López-Lagunas, D. S. Wills, N. Jokerst, M. Brooke
{"title":"Systolic processing architectures using optoelectronic interconnects","authors":"S. Chai, A. López-Lagunas, D. S. Wills, N. Jokerst, M. Brooke","doi":"10.1109/MPPOI.1997.609179","DOIUrl":"https://doi.org/10.1109/MPPOI.1997.609179","url":null,"abstract":"Systolic arrays have traditionally provided efficient, high performance execution for computation intensive applications. Despite the extensive research in systolic arrays, system designers must continually incorporate new technological advances to improve node communications, I/O bandwidth, and programmability. This paper presents optoelectronic interconnect as a communication method for systolic arrays in early image processing applications. Optoelectronic interconnects provide potentially high I/O bandwidth required to maintain high utilization rate for systolic arrays. In addition, optoelectronic interconnects provides a two-dimensional focal-plane topology ideal for systolic image processing systems. This paper introduces two new systolic architectures that incorporate integrated optoelectronics to provide an extremely compact, high performance, highly efficient image processing system. Several important early image processing applications developed for these architectures are also described.","PeriodicalId":406486,"journal":{"name":"Proceedings of the Fourth International Conference on Massively Parallel Processing Using Optical Interconnections","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124618388","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}