Cuauhtémoc R. Aguilera-Galicia, Omar Lonuoria-Gandara, L. Pizano-Escalante
{"title":"Half-Precision Floating-Point Multiplier IP Core Based on 130 Nm CMOS ASIC Technology","authors":"Cuauhtémoc R. Aguilera-Galicia, Omar Lonuoria-Gandara, L. Pizano-Escalante","doi":"10.1109/LATINCOM.2018.8613231","DOIUrl":null,"url":null,"abstract":"Internet of things applications demand reusable modular designs with low-power consumption. Furthermore, many emerging applications, such as image recognition using machine learning, are low-accuracy tolerant. For these applications, the IEEE-754 half-precision arithmetic is becoming a relevant option for low-power, low-computational cost designs. This article presents a half-precision floating-point multiplier. It is implemented on 130 nm CMOS ASIC technology. The proposed multiplier IP core exhibits low-power consumption, small silicon area, and its accuracy is IEEE-754 compliant.","PeriodicalId":332646,"journal":{"name":"2018 IEEE 10th Latin-American Conference on Communications (LATINCOM)","volume":"2 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2018 IEEE 10th Latin-American Conference on Communications (LATINCOM)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/LATINCOM.2018.8613231","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 2
Abstract
Internet of things applications demand reusable modular designs with low-power consumption. Furthermore, many emerging applications, such as image recognition using machine learning, are low-accuracy tolerant. For these applications, the IEEE-754 half-precision arithmetic is becoming a relevant option for low-power, low-computational cost designs. This article presents a half-precision floating-point multiplier. It is implemented on 130 nm CMOS ASIC technology. The proposed multiplier IP core exhibits low-power consumption, small silicon area, and its accuracy is IEEE-754 compliant.