Optical Computing最新文献

{"title":"Optical Artificial Intelligence Based on Semantic Network Architecture","authors":"T. Yatagai","doi":"10.1364/optcomp.1989.mc1","DOIUrl":"https://doi.org/10.1364/optcomp.1989.mc1","url":null,"abstract":"In symbolic processing, associative network approaches show promise for solving difficult artificial intelligence problems. [1,2] Optical associative networks, including holographic[3,4] and matrix-vector multiplication [5] architectures, are one of the most attractive approaches toward large-scale associative processing. Optics provides both 2-D parallel interconnection ability between modules and parallel-computing mechanisms for parallel association algorithm. A hybrid optical inference architecture has been proposed. [6] Recently optical architectures for learning and self-organizing neural network are discussed.[7,8]","PeriodicalId":302010,"journal":{"name":"Optical Computing","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132403712","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":"Electronic vs. Optical Implementations of Neural Networks*","authors":"J. Sage","doi":"10.1364/optcomp.1989.ma2","DOIUrl":"https://doi.org/10.1364/optcomp.1989.ma2","url":null,"abstract":"This paper will address for the optical community the relative advantages and limitations of electronic neural network implementations in contrast to optical implementations. Its intent is by no means to be adversarial. It is hardly necessary to say that today electronics has an edge over optics in this field. The aim of the paper will be to help indicate the areas where electronic and optical implementations can each make their most important contributions and the areas in which advances in technology, particularly in optical technology, will be required.","PeriodicalId":302010,"journal":{"name":"Optical Computing","volume":"116 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133544993","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":"High-Density 300-Gbps/cm2 Parallel Free-Space Optical Interconnection Design Considerations","authors":"D. Tsang","doi":"10.1364/optcomp.1995.othb2","DOIUrl":"https://doi.org/10.1364/optcomp.1995.othb2","url":null,"abstract":"A high-density, high-throughput 300-Gbps/cm2 parallel free-space optical interconnection has been designed and demonstrated. The impact of component technology choices and optical, electrical, and mechanical issues will be discussed within the context of this prototype system with implications for future systems. Many of the design considerations for this prototype are common to other optical interconnection and processing systems based on arrays of components. The prototype, a linear array parallel free-space optical interconnection with up to twenty optical data paths, operated at a rate of up to 2.8 Gbps per optical data path with a delay or latency of 200 ps.","PeriodicalId":302010,"journal":{"name":"Optical Computing","volume":"219 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133580590","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":"The Thermal Nonlinear Microcavity and Optical Computing","authors":"C. Godsalve, E. Abraham","doi":"10.1364/optcomp.1989.tui24","DOIUrl":"https://doi.org/10.1364/optcomp.1989.tui24","url":null,"abstract":"Thin film Fabry-Perot etalons which have a temperature dependent refractive index exhibit bistability or gain at room temperature (e.g. ZnSe) and at optical frequencies. These features make them candidates for digital optical computing. An N x N array of elements can be generated in a single filter by an array of laser beams. As a result, thermal crosstalk develops which is long range and only a few elements per cm2 on such a filter can operate independently [1,2]. However if each filter is mounted on its own separate 'turret', crosstalk can be reduced to the extent that 104 microcavities (or pixels) can operate independently per cm2 [3,4].","PeriodicalId":302010,"journal":{"name":"Optical Computing","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133701176","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":"Implementation of a 16-channel Sorting Module","authors":"D. Baillie, F. Tooley, S. Prince, N. L. Grant, J. Dines, M. Desmulliez, M. Taghizadeh","doi":"10.1364/optcomp.1995.oma2","DOIUrl":"https://doi.org/10.1364/optcomp.1995.oma2","url":null,"abstract":"This paper will present experimental details of a sorting module demonstration system. The sorting module which is currently under construction is shown as a functional schematic in figure 1. Figure 2 is a photograph of the optics. The system implements the bitonic sort based on Batcher’s algorithm implemented with a perfect shuffle. A re-circulating rather than pipelined arrangement is used to minimise hardware requirements to 2 smart pixel chips;\u0000 • a sorting node array (self-routing exchange/bypass nodes), and\u0000 • shift register array which acts as the input/output interface.","PeriodicalId":302010,"journal":{"name":"Optical Computing","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133343918","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":"High-Speed Database Processing on an Optical Content-Addressable Parallel Processor(OCAPP)","authors":"A. Louri, J. A. Hatch","doi":"10.1364/optcomp.1993.otua.10","DOIUrl":"https://doi.org/10.1364/optcomp.1993.otua.10","url":null,"abstract":"Recently, an efficient optical architecture, known as Optical Content-Addressable Parallel Processor(OCAPP), has been introduced for the efficient implementation of symbolic computing tasks [1].","PeriodicalId":302010,"journal":{"name":"Optical Computing","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133437826","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":"Massively Parallel Processing with Optical Interconnections: What Can Be, Should Be and Must Not Be Done By Optics","authors":"E. Schenfeld","doi":"10.1364/optcomp.1995.omb1","DOIUrl":"https://doi.org/10.1364/optcomp.1995.omb1","url":null,"abstract":"What is wrong about Optical Computing is the implied search for “general purpose computing”. We think that such an attempt has little chance to result in a practical system for, at least, the next ten years. The main reason is the economical justification. What such an “optical computing” system may offer has to be compared with the value of the application and the alternatives (electronics). On the other hand, communication in general is an area where optics has proved to be a real blessing. Long distance communication is most economically done today using optical fibers. We think that another realistic search for good optical applications should now be done for shorter distances. A possible good direction may be the communication needs of Massively Parallel Processing (MPP) systems. In such a system, large number (10’s of thousands) of Processing Elements (PEs) are to be interconnected. A PE can be seen as made of a high-end single chip CPU available today, with memory and communication circuits. We do not view the other possible meaning of MPP, namely processing and interconnections at the single gate or device level, as practical to consider. This paper describes the views of the author from the computer architecture’s standpoint, with the hope to serve as a pointer to the “Optical Computing” community. Although much has been done in the area of optical communication technology for the past 10-20 years, and many optical network experimental systems have been proposed, it seems that optics has not yet found its expected place as the interconnection technology of choice for MPP systems. In this paper we try to suggest some possible reasons preventing the common use of optical interconnections in MPP systems, in a hope to focus attention on what really needs to be done to advance the field. We would suggest focusing on searching for a processing-less solution rather than trying to mimic the existing thinking of electronic networks. We outline several key principles essential to follow to reach realistic and economical solutions of optical interconnections for MPP systems. An example of using such principles for an MPP, free-space network is presented in [1].","PeriodicalId":302010,"journal":{"name":"Optical Computing","volume":"141 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132649634","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-Pumped Optical Neural Networks","authors":"Y. Owechko","doi":"10.1364/optcomp.1989.md4","DOIUrl":"https://doi.org/10.1364/optcomp.1989.md4","url":null,"abstract":"Neural network models for artificial intelligence offer an approach fundamentally different from conventional symbolic approaches, but the merits of the two paradigms cannot be fairly compared until neural network models with large numbers of ”neurons” are implemented. Despite the attractiveness of neural networks for computing applications which involve adaptation and learning, most of the published demonstrations of neural network technology have involved relatively small numbers of ”neurons”. One reason for this is the poor match between conventional electronic serial or coarse-grained multiple-processor computers and the massive parallelism and communication requirements of neural network models. The self-pumped optical neural network (SPONN) described here is a fine-grained optical architecture which features massive parallelism and a much greater degree of interconnectivity than bus-oriented or hypercube electronic architectures. SPONN is potentially capable of implementing neural networks consisting of 105-106 neurons with 109-1010 interconnections. The mapping of neural network models onto the architecture occurs naturally without the need for multiplexing neurons or dealing with contention, routing, and communication bottleneck problems. This simplifies the programming involved compared to electronic implementations.","PeriodicalId":302010,"journal":{"name":"Optical Computing","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132658588","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":"Tantalus and Optical Computing","authors":"W. Cathey","doi":"10.1364/optcomp.1989.tuf1","DOIUrl":"https://doi.org/10.1364/optcomp.1989.tuf1","url":null,"abstract":"The ancient Greek, Tantalus, had a problem that is similar to that of some optical computing and signal processor researchers. As a punishment by the Greek gods, Tantalus was placed in a lake with water up to his waist. Fruit was on branches just above his head. However, when he leaned over to drink, or reached up to eat, the water receded just beyond reach and the fruit evaded his grasp. Hence, he was doomed to never obtaining what seemed so close. I do not know what deed was done by the optics researchers that deserves similar punishment (I have some ideas.), but some of the anticipated results and applications of optics to computing seem to be always just barely beyond reach.","PeriodicalId":302010,"journal":{"name":"Optical Computing","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132677931","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":"Routing Algorithm for a Circuit-switched Optical Extended Generalized Shuffle Network","authors":"C. Waterson, B. K. Jenkins","doi":"10.1364/optcomp.1995.omb5","DOIUrl":"https://doi.org/10.1364/optcomp.1995.omb5","url":null,"abstract":"Two key difficulties in the implementation and use of multistage interconnection networks have been the complexity of the network hardware and the complexity of the routing algorithm. This has been particularly evident in MIMD computing environments, when the network needs to support arbitrary interconnection pattern requests, and when no requests are to be buffered or postponed. Under the premise that the use of optics can at least partially alleviate the network hardware complexity issue, we consider in this paper the routing algorithm complexity needed to control such an optical network. We present a routing algorithm designed with the goal of minimal time complexity.","PeriodicalId":302010,"journal":{"name":"Optical Computing","volume":"872 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131899176","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}