{"title":"Design of a 3-D stacked floating-point Goldschmidt divider","authors":"J. Tada, Ryusuke Egawa, Hiroaki Kobayashi","doi":"10.1109/3DIC.2015.7334598","DOIUrl":null,"url":null,"abstract":"In the design of 3-D stacked floating-point units, a partitioning method affects the performance and the power consumption. To realize a high-performance and low-power 3-D stacked floating point divider, this paper proposes a circuit partitioning method for the Goldschmidt divider. The proposed partitioning method equalizes the sizes of silicon layers and reduces the number of vertical interconnects. Experimental results show the 3-D stacked Goldschmidt divider which is designed based on the proposed partitioning method achieves an 8.1% critical path delay reduction and a 6.8% average power reduction compared to the 2-D implementation.","PeriodicalId":253726,"journal":{"name":"2015 International 3D Systems Integration Conference (3DIC)","volume":"15 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2015-11-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2015 International 3D Systems Integration Conference (3DIC)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/3DIC.2015.7334598","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
In the design of 3-D stacked floating-point units, a partitioning method affects the performance and the power consumption. To realize a high-performance and low-power 3-D stacked floating point divider, this paper proposes a circuit partitioning method for the Goldschmidt divider. The proposed partitioning method equalizes the sizes of silicon layers and reduces the number of vertical interconnects. Experimental results show the 3-D stacked Goldschmidt divider which is designed based on the proposed partitioning method achieves an 8.1% critical path delay reduction and a 6.8% average power reduction compared to the 2-D implementation.