Proceedings. 6th International Conference on Parallel Interconnects (PI'99) (Formerly Known as MPPOI)最新文献

PAROLI(R)-a parallel optical link with 15 Gbit/s throughput in a 12-channel wide interconnection PAROLI(R)-在12通道宽互连中具有15gbit /s吞吐量的并行光链路

Proceedings. 6th International Conference on Parallel Interconnects (PI'99) (Formerly Known as MPPOI) Pub Date : 1999-10-17 DOI: 10.1109/PI.1999.806411

D. Kuhl, K. Drogemuller, J. Blank, M. Ehlert, T. Kraeker, J. Hohn, D. Klix, V. Plickert, L. Melchior, P. Hildebrandt, M. Heinemann, A. Beier, L. Leininger, H.-D. Wolf, T. Wipiejewski, R. Engel

引用次数: 4

A multi-layer-perceptron neural network hardware based on 3D massively parallel optoelectronic circuits 一种基于三维大规模并行光电电路的多层感知器神经网络硬件

Proceedings. 6th International Conference on Parallel Interconnects (PI'99) (Formerly Known as MPPOI) Pub Date : 1999-10-17 DOI: 10.1109/PI.1999.806397

K. D. Maier, C. Beckstein, Reinhard Blickhan, Werner Erhard, Dietmar Fey

引用次数: 13

3-D optical networks of directional couplers and Mach-Zehnder 定向耦合器和Mach-Zehnder的三维光网络

Proceedings. 6th International Conference on Parallel Interconnects (PI'99) (Formerly Known as MPPOI) Pub Date : 1999-10-17 DOI: 10.1109/PI.1999.806415

J. Giglmayr

引用次数: 0

Matrix multiplication on the OTIS-Mesh optoelectronic computer OTIS-Mesh光电计算机上的矩阵乘法

Proceedings. 6th International Conference on Parallel Interconnects (PI'99) (Formerly Known as MPPOI) Pub Date : 1999-10-17 DOI: 10.1109/PI.1999.806404

Chih-Fang Wang, S. Sahni

引用次数: 84

Parallel optical interconnects for high performance printed circuit boards 用于高性能印刷电路板的并行光学互连

Proceedings. 6th International Conference on Parallel Interconnects (PI'99) (Formerly Known as MPPOI) Pub Date : 1999-10-17 DOI: 10.1109/PI.1999.806410

E. Griese

{"title":"Parallel optical interconnects for high performance printed circuit boards","authors":"E. Griese","doi":"10.1109/PI.1999.806410","DOIUrl":"https://doi.org/10.1109/PI.1999.806410","url":null,"abstract":"A novel and innovative interconnection technology for printed circuit board application is presented which is able to meet the high performance requirements of future electronic equipment while at the same time improving the electromagnetic compatibility (EMC) significantly. This technology will have a far-reaching compatibility with the existing printed circuit board technology, which means that the design and manufacturing processes of the electrical part do not need significant modifications. After a short description of the most important basic technologies and results for its realization, the paper focuses on the design and modeling of parallel electrical/optical interconnection systems. Transient analysis is addressed in order to provide an efficient analysis methodology as well as algorithms for timing and signal integrity prediction necessary for designing and manufacturing high-speed electronic systems of high quality. The developed overall modeling strategy is explained and first available results are presented. Besides a time domain transmission line model for parallel optical multimode structures considering delay, losses, dispersion, and crosstalk the corresponding modeling approaches for laser- and photo-diodes are presented.","PeriodicalId":157032,"journal":{"name":"Proceedings. 6th International Conference on Parallel Interconnects (PI'99) (Formerly Known as MPPOI)","volume":"144 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1999-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132400468","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}

引用次数: 13

Ultra-fast low-latent superconductor packet switching networks for petaflops computing 用于千万亿次浮点运算的超高速低潜超导分组交换网络

Proceedings. 6th International Conference on Parallel Interconnects (PI'99) (Formerly Known as MPPOI) Pub Date : 1999-10-17 DOI: 10.1109/PI.1999.806396

D. Zinoviev, L. Wittie

引用次数: 0

The PHOTOBUS smart pixel interconnection system for symmetric multiprocessing using workstation clusters PHOTOBUS智能像素互连系统，使用工作站集群进行对称多处理

Proceedings. 6th International Conference on Parallel Interconnects (PI'99) (Formerly Known as MPPOI) Pub Date : 1999-10-17 DOI: 10.1109/PI.1999.806401

P. Lukowicz

引用次数: 5

Disk write caching with an optical network 基于光网络的磁盘写缓存

Proceedings. 6th International Conference on Parallel Interconnects (PI'99) (Formerly Known as MPPOI) Pub Date : 1999-10-17 DOI: 10.1109/PI.1999.806414

E. V. Carrera, Ricardo Bianchini

{"title":"Disk write caching with an optical network","authors":"E. V. Carrera, Ricardo Bianchini","doi":"10.1109/PI.1999.806414","DOIUrl":"https://doi.org/10.1109/PI.1999.806414","url":null,"abstract":"In this paper we propose a simple extension to the optical network of a scalable multiprocessor that optimizes page swap-outs significantly. More specifically, we propose to extend the network with an optical ring that not only transfers swapped-out pages between the local memories and the disks of the multiprocessor but also acts as a system-wide write cache for these pages. This extended optical network has several performance benefits: it provides a staging area where swapped-out pages can reside until the disk is free; it increases the possibility of combining several writes to disk; and it acts as a victim cache for pages that are swapped out and subsequently accessed by the same or a different processor. In order to evaluate the extent to which these benefits affect performance, we use detailed execution-driven simulations of several out-of-core parallel applications running on an 8-node scalable multiprocessor. Our results demonstrate that our optical ring provides consistent performance improvements, coming mostly from faster page swap-outs and victim caching. Based on these results and on our parameter space study, our main conclusion is that our optical ring is highly efficient under several architectural assumptions and for most out-of-core parallel applications.","PeriodicalId":157032,"journal":{"name":"Proceedings. 6th International Conference on Parallel Interconnects (PI'99) (Formerly Known as MPPOI)","volume":"34 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1999-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121895162","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}

引用次数: 1

Quantifying the impact of optical interconnect latency on the performance of optoelectronic FPGAs 量化光互连延迟对光电fpga性能的影响

Proceedings. 6th International Conference on Parallel Interconnects (PI'99) (Formerly Known as MPPOI) Pub Date : 1999-10-17 DOI: 10.1109/PI.1999.806399

J. Dambre, H. V. Van Marck, J. V. Van Campenhout

引用次数: 11

From HiPPI-800 to HiPPI-6400: A changing of the guard and gateway to the future 从HiPPI-800到HiPPI-6400:改变守卫和通往未来的大门

Proceedings. 6th International Conference on Parallel Interconnects (PI'99) (Formerly Known as MPPOI) Pub Date : 1999-10-17 DOI: 10.1109/PI.1999.806412

D. Tolmie, T. Boorman, A. DuBois, D. Dubois, Wu-chun Feng, I. Philp

{"title":"From HiPPI-800 to HiPPI-6400: A changing of the guard and gateway to the future","authors":"D. Tolmie, T. Boorman, A. DuBois, D. Dubois, Wu-chun Feng, I. Philp","doi":"10.1109/PI.1999.806412","DOIUrl":"https://doi.org/10.1109/PI.1999.806412","url":null,"abstract":"HiPPI-6400, a high-performance parallel interface running at 6400 Mb/s (800 MB/s), is a networking technology targeted for deployment in a local-area network (LAN) or system-area network (SAN). It is a low-latency, high-bandwidth switch that has the added features of providing flow control and error detection and retransmission in hardware, thus freeing network software from having to implement these functions. Due to the very low overhead of the HiPPI-6400 physical layer and the use of separate control lines, user data rates of HiPPI-6400 are literally 6400 Mb/s or 6.4 Gb/s (eight times the usable bandwidth of Gigabit Ethernet) with an achievable bit-error rate of less than 10/sup 25/. An initial 32-port HiPPI-6400 prototype, running an OS-bypass network protocol called ST between two 32-node SGI Origin 2000s, produced one-way latencies of 7 /spl mu/s and sustained data rates of 3.6 Gb/s (unidirectional) and 6.4 Gb/s (bidirectional). The HiPPI-6400 bandwidth in both cases was limited by the memory architecture of the SGI Origin 2000, not by the network. When running an OS-based protocol such as TCP, the unidirectional bandwidth was 2.2 Gb/s.","PeriodicalId":157032,"journal":{"name":"Proceedings. 6th International Conference on Parallel Interconnects (PI'99) (Formerly Known as MPPOI)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1999-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122035726","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}

引用次数: 10