Pub Date : 2019-02-01DOI: 10.1109/HiPINEB.2019.00008
Ryland Curtsinger, David P. Bunde
Dragonfly is a topology for high-performance computer systems designed to exploit technology trends and meet challenging system constraints, particularly on power. In a Dragonfly system, compute nodes are attached to switches, the switches are organized into groups, and the network is organized as a two-level clique, with an edge between every switch in a group and an edge between every pair of groups. This means that every pair of switches is separated by at most three hops, one within a source group, one from the source group to the destination group, and one within the destination group. Routing using paths of this form is typically called "minimal routing". In this paper, we show that the resulting paths are not always the shortest possible. We then propose a new class of paths that can be used without additional networking hardware and count its members that are shorter than or of equal length to these "minimal paths".
{"title":"Shortest paths in Dragonfly systems","authors":"Ryland Curtsinger, David P. Bunde","doi":"10.1109/HiPINEB.2019.00008","DOIUrl":"https://doi.org/10.1109/HiPINEB.2019.00008","url":null,"abstract":"Dragonfly is a topology for high-performance computer systems designed to exploit technology trends and meet challenging system constraints, particularly on power. In a Dragonfly system, compute nodes are attached to switches, the switches are organized into groups, and the network is organized as a two-level clique, with an edge between every switch in a group and an edge between every pair of groups. This means that every pair of switches is separated by at most three hops, one within a source group, one from the source group to the destination group, and one within the destination group. Routing using paths of this form is typically called \"minimal routing\". In this paper, we show that the resulting paths are not always the shortest possible. We then propose a new class of paths that can be used without additional networking hardware and count its members that are shorter than or of equal length to these \"minimal paths\".","PeriodicalId":371832,"journal":{"name":"2019 International Workshop of High-Perfomance Interconnection Networks in the Exascale and Big-Data Era (HiPNEB)","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129483718","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 : 2019-02-01DOI: 10.1109/HiPINEB.2019.00009
F. Zahn, P. Yébenes, J. Escudero-Sahuquillo, P. García, H. Fröning
In post-Dennard scaling energy becomes more and more important. While most components in data-center and supercomputer become increasingly energy-proportional, this trend seems to pass on interconnection networks. Although previous studies have shown huge potential for saving energy in interconnects, the associated performance decrease seems to be obstructive. An increase of execution time can be caused by a decreased bandwidth as well as by transition times which links need to reconfigure and are not able to transmit data. This leads to more contention on the network than usually interconnects have to deal with. Congestion management is used in similar situations to limit the impact of these contentions only to single links and avoiding them to congest the entire network. Therefore, we propose combining energy saving policies and congestion management queueing schemes in order to maintain performance while saving energy. For synthetic hotspot traffic, which we use to stress the network, this combination shows promising results for multiple topologies. In 3D torus, k-ary n-tree, and dragonfly this combination provides a more than 50% lower latency and increases energy efficiency by more than 50% compared to the baseline. Although both techniques aim for fundamental different goals, non of the investigated configurations seems to suffer any disadvantages from their combination.
{"title":"Effects of Congestion Management on Energy Saving Techniques in Interconnection Networks","authors":"F. Zahn, P. Yébenes, J. Escudero-Sahuquillo, P. García, H. Fröning","doi":"10.1109/HiPINEB.2019.00009","DOIUrl":"https://doi.org/10.1109/HiPINEB.2019.00009","url":null,"abstract":"In post-Dennard scaling energy becomes more and more important. While most components in data-center and supercomputer become increasingly energy-proportional, this trend seems to pass on interconnection networks. Although previous studies have shown huge potential for saving energy in interconnects, the associated performance decrease seems to be obstructive. An increase of execution time can be caused by a decreased bandwidth as well as by transition times which links need to reconfigure and are not able to transmit data. This leads to more contention on the network than usually interconnects have to deal with. Congestion management is used in similar situations to limit the impact of these contentions only to single links and avoiding them to congest the entire network. Therefore, we propose combining energy saving policies and congestion management queueing schemes in order to maintain performance while saving energy. For synthetic hotspot traffic, which we use to stress the network, this combination shows promising results for multiple topologies. In 3D torus, k-ary n-tree, and dragonfly this combination provides a more than 50% lower latency and increases energy efficiency by more than 50% compared to the baseline. Although both techniques aim for fundamental different goals, non of the investigated configurations seems to suffer any disadvantages from their combination.","PeriodicalId":371832,"journal":{"name":"2019 International Workshop of High-Perfomance Interconnection Networks in the Exascale and Big-Data Era (HiPNEB)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129565303","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 : 2019-02-01DOI: 10.1109/hipineb.2019.00005
{"title":"HiPINEB 2019 Welcome Message","authors":"","doi":"10.1109/hipineb.2019.00005","DOIUrl":"https://doi.org/10.1109/hipineb.2019.00005","url":null,"abstract":"","PeriodicalId":371832,"journal":{"name":"2019 International Workshop of High-Perfomance Interconnection Networks in the Exascale and Big-Data Era (HiPNEB)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133546784","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 : 2019-02-01DOI: 10.1109/hipineb.2019.00007
{"title":"Keynote by Dhabaleswar K. (DK) Panda","authors":"","doi":"10.1109/hipineb.2019.00007","DOIUrl":"https://doi.org/10.1109/hipineb.2019.00007","url":null,"abstract":"","PeriodicalId":371832,"journal":{"name":"2019 International Workshop of High-Perfomance Interconnection Networks in the Exascale and Big-Data Era (HiPNEB)","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115011418","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：