Pub Date : 1998-06-15DOI: 10.1109/MPPOI.1998.682139
M. Jonsson
In this paper, we propose a control-channel based ring network built up of fiber-ribbon point-to-point links. One of the fibers in each link forms part of the control-channel ring, over which medium access control information is sent immediately before data transmissions. This increases performance of the network. High throughputs can be achieved in the network due to pipelining, i.e., several packets can be traveling through the network simultaneously but in different segments of the ring. The network can meet high performance demands in, e.g., massively parallel signal processing systems, which is shown by example in the paper. Also, real-time demands can be met using slot reserving. The network, called CCFPR (Control-Channel based Fiber-ribbon Pipeline Ring), can be built today, using fiber-optic off-the-shelf components, and a prototype is currently under development. The increasingly good price/performance ratio for fiber-ribbon links indicates a great success potential for the proposed kind of network.
{"title":"Control-channel based fiber-ribbon pipeline ring network","authors":"M. Jonsson","doi":"10.1109/MPPOI.1998.682139","DOIUrl":"https://doi.org/10.1109/MPPOI.1998.682139","url":null,"abstract":"In this paper, we propose a control-channel based ring network built up of fiber-ribbon point-to-point links. One of the fibers in each link forms part of the control-channel ring, over which medium access control information is sent immediately before data transmissions. This increases performance of the network. High throughputs can be achieved in the network due to pipelining, i.e., several packets can be traveling through the network simultaneously but in different segments of the ring. The network can meet high performance demands in, e.g., massively parallel signal processing systems, which is shown by example in the paper. Also, real-time demands can be met using slot reserving. The network, called CCFPR (Control-Channel based Fiber-ribbon Pipeline Ring), can be built today, using fiber-optic off-the-shelf components, and a prototype is currently under development. The increasingly good price/performance ratio for fiber-ribbon links indicates a great success potential for the proposed kind of network.","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":"129579064","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.1998.682128
T. Sterling
This talk examines Petaflops scale computing as understood from the HTMT architecture and other design studies with a focus on their implications for optical communications. It is shown that for certain parts of the total communication budget, only optical technologies are likely to serve. The Data Vortex being developed by Princeton University is briefly described as an example of a possible network class that can meet the challenges described for the intra-memory fabric. This talk concludes with a set of near term challenges in the area of practical fabrication and integration of electrooptic technology necessary for meeting to needs of Petaflops scale computing systems.
{"title":"Opportunities for optical communications in Petaflops-scale computing","authors":"T. Sterling","doi":"10.1109/MPPOI.1998.682128","DOIUrl":"https://doi.org/10.1109/MPPOI.1998.682128","url":null,"abstract":"This talk examines Petaflops scale computing as understood from the HTMT architecture and other design studies with a focus on their implications for optical communications. It is shown that for certain parts of the total communication budget, only optical technologies are likely to serve. The Data Vortex being developed by Princeton University is briefly described as an example of a possible network class that can meet the challenges described for the intra-memory fabric. This talk concludes with a set of near term challenges in the area of practical fabrication and integration of electrooptic technology necessary for meeting to needs of Petaflops scale computing systems.","PeriodicalId":248808,"journal":{"name":"Proceedings. Fifth International Conference on Massively Parallel Processing (Cat. No.98EX182)","volume":"9 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":"134481427","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.1998.682122
J. Jahns
Free-space optics can help alleviate the communications problems of electronic computers by its large spatial and temporal bandwidth. For compact packaging, the concept of planar-integrated free-space optics is used. Here, we give an overview of the design, fabrication and thermal management of planar optical interconnects.
{"title":"Integrated free-space optical interconnects for chip-to-chip communications","authors":"J. Jahns","doi":"10.1109/MPPOI.1998.682122","DOIUrl":"https://doi.org/10.1109/MPPOI.1998.682122","url":null,"abstract":"Free-space optics can help alleviate the communications problems of electronic computers by its large spatial and temporal bandwidth. For compact packaging, the concept of planar-integrated free-space optics is used. Here, we give an overview of the design, fabrication and thermal management of planar optical interconnects.","PeriodicalId":248808,"journal":{"name":"Proceedings. Fifth International Conference on Massively Parallel Processing (Cat. No.98EX182)","volume":"33 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":"122671633","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.1998.682126
B. Webb, A. Louri
Building large crossbar switches for low-latency, high-bandwidth interprocessor interconnects in multiprocessor systems is a difficult task because of the many practical limitations of digital electronics. This paper presents a first order design for a crossbar switch which utilizers wavelength tunable vertical-cavity surface-emitting laser (VCSEL) technology and free space optics. Free space optics can be utilized to construct a crossbar switch without the wiring complexity normally associated with digital electronic crossbar. Wavelength tunable VCSELs can be utilized to reduce the required number of components per node to as few as one wavelength tunable VCSEL and one fixed frequency receiver, thereby reducing the overall cost. By utilizing space division multiplexing, one can further increase the size of the optical crossbar switch by adding additional wavelength tunable VCSELs at each node, greatly increasing the saleability of the overall system. An example system configuration containing 256 channels is presented.
{"title":"A free space optical crossbar switch using wavelength division multiplexing and vertical-cavity surface-emitting lasers","authors":"B. Webb, A. Louri","doi":"10.1109/MPPOI.1998.682126","DOIUrl":"https://doi.org/10.1109/MPPOI.1998.682126","url":null,"abstract":"Building large crossbar switches for low-latency, high-bandwidth interprocessor interconnects in multiprocessor systems is a difficult task because of the many practical limitations of digital electronics. This paper presents a first order design for a crossbar switch which utilizers wavelength tunable vertical-cavity surface-emitting laser (VCSEL) technology and free space optics. Free space optics can be utilized to construct a crossbar switch without the wiring complexity normally associated with digital electronic crossbar. Wavelength tunable VCSELs can be utilized to reduce the required number of components per node to as few as one wavelength tunable VCSEL and one fixed frequency receiver, thereby reducing the overall cost. By utilizing space division multiplexing, one can further increase the size of the optical crossbar switch by adding additional wavelength tunable VCSELs at each node, greatly increasing the saleability of the overall system. An example system configuration containing 256 channels is presented.","PeriodicalId":248808,"journal":{"name":"Proceedings. Fifth International Conference on Massively Parallel Processing (Cat. No.98EX182)","volume":"180 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":"116440806","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.1998.682136
J. P. Lang, Emmanouel Varvarigos, V. Sharma
We present a new analysis of wavelength translation in regular all-optical WDM networks, that is simple, computationally inexpensive, and accurate for both low and high network loads. In a network with k wavelengths per link, we model the output link by an auxiliary M/M/k/k queueing system. We then obtain a closed-form expression for the probability P/sub succ/ that a session arriving at a node at a random time successfully establishes a connection from its source node to its destination node. Unlike previous analyses, which use the link independence blocking assumption, we account for the dependence between the acquisition of wavelengths on successive links of the session's path. Based on the success probability, we show that the throughput per wavelength increases superlinearly (as expected) as we increase the number of wavelengths per link; however the extent of this superlinear increase in throughput saturates rather quickly. This suggests some interesting possibilities for network provisioning in an all-optical network. We verify the accuracy of our analysis via simulations for the torus and hypercube networks.
{"title":"A new analysis for wavelength translation in regular WDM networks","authors":"J. P. Lang, Emmanouel Varvarigos, V. Sharma","doi":"10.1109/MPPOI.1998.682136","DOIUrl":"https://doi.org/10.1109/MPPOI.1998.682136","url":null,"abstract":"We present a new analysis of wavelength translation in regular all-optical WDM networks, that is simple, computationally inexpensive, and accurate for both low and high network loads. In a network with k wavelengths per link, we model the output link by an auxiliary M/M/k/k queueing system. We then obtain a closed-form expression for the probability P/sub succ/ that a session arriving at a node at a random time successfully establishes a connection from its source node to its destination node. Unlike previous analyses, which use the link independence blocking assumption, we account for the dependence between the acquisition of wavelengths on successive links of the session's path. Based on the success probability, we show that the throughput per wavelength increases superlinearly (as expected) as we increase the number of wavelengths per link; however the extent of this superlinear increase in throughput saturates rather quickly. This suggests some interesting possibilities for network provisioning in an all-optical network. We verify the accuracy of our analysis via simulations for the torus and hypercube networks.","PeriodicalId":248808,"journal":{"name":"Proceedings. Fifth International Conference on Massively Parallel Processing (Cat. No.98EX182)","volume":"8 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":"127489671","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.1998.682125
Chih-Hao Chen, B. Hoanca, Charles B. Kuznia, Alexander A. Sawchuk, Jen-Ming Wu
We present a smart pixel device for networking and parallel processing based on the use two-dimensional arrays optical channels between VLSI chip planes.
我们提出了一种基于在VLSI芯片平面之间使用二维阵列光通道的智能像素设备，用于网络和并行处理。
{"title":"TRANslucent smart pixel ARray (TRANSPAR) chips for high throughput networks and SIMD signal processing","authors":"Chih-Hao Chen, B. Hoanca, Charles B. Kuznia, Alexander A. Sawchuk, Jen-Ming Wu","doi":"10.1109/MPPOI.1998.682125","DOIUrl":"https://doi.org/10.1109/MPPOI.1998.682125","url":null,"abstract":"We present a smart pixel device for networking and parallel processing based on the use two-dimensional arrays optical channels between VLSI chip planes.","PeriodicalId":248808,"journal":{"name":"Proceedings. Fifth International Conference on Massively Parallel Processing (Cat. No.98EX182)","volume":"164 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":"127426300","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 : 1998-04-23DOI: 10.1109/MPPOI.1998.682119
L. Coldren, E. Hegblom, Y. Akulova, J. Ko, E. Strzelecka, S. Hu
Continuing increases in efficiency, uniformity, and yield for low-threshold dielectrically Apertured Vertical-Cavity Surface-Emitting Lasers (VCSELs) suggest that this new generation of sources may be ready for insertion into practical parallel interconnect systems. This paper will review the recent evolution of these devices, pointing out key enabling advances and potential roadblocks yet to be addressed. Included will be advances that have led to record low optical losses as well as record high wall-plug efficiency at powers of a few hundred microwatts, desirable for massively parallel optical interconnects. The use of engineered oxide apertures is a key element in these cases. Experimental results will also include recent free-space and WDM fiber interconnects. Remaining issues to be addressed include some sort of lateral carrier confinement, such as a buried-heterostructure, to reduce carrier losses as devices are scaled to small lateral dimensions.
{"title":"Vertical-cavity lasers for parallel optical interconnects","authors":"L. Coldren, E. Hegblom, Y. Akulova, J. Ko, E. Strzelecka, S. Hu","doi":"10.1109/MPPOI.1998.682119","DOIUrl":"https://doi.org/10.1109/MPPOI.1998.682119","url":null,"abstract":"Continuing increases in efficiency, uniformity, and yield for low-threshold dielectrically Apertured Vertical-Cavity Surface-Emitting Lasers (VCSELs) suggest that this new generation of sources may be ready for insertion into practical parallel interconnect systems. This paper will review the recent evolution of these devices, pointing out key enabling advances and potential roadblocks yet to be addressed. Included will be advances that have led to record low optical losses as well as record high wall-plug efficiency at powers of a few hundred microwatts, desirable for massively parallel optical interconnects. The use of engineered oxide apertures is a key element in these cases. Experimental results will also include recent free-space and WDM fiber interconnects. Remaining issues to be addressed include some sort of lateral carrier confinement, such as a buried-heterostructure, to reduce carrier losses as devices are scaled to small lateral dimensions.","PeriodicalId":248808,"journal":{"name":"Proceedings. Fifth International Conference on Massively Parallel Processing (Cat. No.98EX182)","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-04-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129880505","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：