Junyoung Park, Injoon Hong, Gyeonghoon Kim, Youchang Kim, K. Lee, Seongwook Park, Kyeongryeol Bong, H. Yoo
{"title":"具有内容感知的细粒度任务调度的多粒度并行对象识别处理器","authors":"Junyoung Park, Injoon Hong, Gyeonghoon Kim, Youchang Kim, K. Lee, Seongwook Park, Kyeongryeol Bong, H. Yoo","doi":"10.1109/CoolChips.2013.6547917","DOIUrl":null,"url":null,"abstract":"Multiple granularity parallel core architecture is proposed to accelerate object recognition with low area and energy consumption. By adopting task-level optimized cores with different parallelism and complexity, the proposed processor achieves real-time object recognition with 271.4 GOPS peak performance. In addition, content-aware fine-grained task scheduling is proposed to enable low power real-time object recognition on 30fps 720p HD video streams. As a result, the object recognition processor achieves 9.4nJ/pixel energy efficiency and 25.8 GOPS/W·mm2 power-area efficiency in O.13um CMOS technology.","PeriodicalId":340576,"journal":{"name":"2013 IEEE COOL Chips XVI","volume":"9 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2013-04-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":"{\"title\":\"A multi-granularity parallelism object recognition processor with content-aware fine-grained task scheduling\",\"authors\":\"Junyoung Park, Injoon Hong, Gyeonghoon Kim, Youchang Kim, K. Lee, Seongwook Park, Kyeongryeol Bong, H. Yoo\",\"doi\":\"10.1109/CoolChips.2013.6547917\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Multiple granularity parallel core architecture is proposed to accelerate object recognition with low area and energy consumption. By adopting task-level optimized cores with different parallelism and complexity, the proposed processor achieves real-time object recognition with 271.4 GOPS peak performance. In addition, content-aware fine-grained task scheduling is proposed to enable low power real-time object recognition on 30fps 720p HD video streams. As a result, the object recognition processor achieves 9.4nJ/pixel energy efficiency and 25.8 GOPS/W·mm2 power-area efficiency in O.13um CMOS technology.\",\"PeriodicalId\":340576,\"journal\":{\"name\":\"2013 IEEE COOL Chips XVI\",\"volume\":\"9 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2013-04-17\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"1\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2013 IEEE COOL Chips XVI\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/CoolChips.2013.6547917\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2013 IEEE COOL Chips XVI","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/CoolChips.2013.6547917","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 1
摘要
提出了多粒度并行核架构，以提高目标识别的速度，同时具有低面积和低能耗。该处理器采用不同并行度和复杂度的任务级优化内核，以271.4 GOPS的峰值性能实现实时目标识别。此外，还提出了内容感知的细粒度任务调度，以实现30fps 720p高清视频流下的低功耗实时目标识别。因此，该目标识别处理器在0.13 um CMOS技术下实现了9.4nJ/像素的能量效率和25.8 GOPS/W·mm2的功率面积效率。
A multi-granularity parallelism object recognition processor with content-aware fine-grained task scheduling
Multiple granularity parallel core architecture is proposed to accelerate object recognition with low area and energy consumption. By adopting task-level optimized cores with different parallelism and complexity, the proposed processor achieves real-time object recognition with 271.4 GOPS peak performance. In addition, content-aware fine-grained task scheduling is proposed to enable low power real-time object recognition on 30fps 720p HD video streams. As a result, the object recognition processor achieves 9.4nJ/pixel energy efficiency and 25.8 GOPS/W·mm2 power-area efficiency in O.13um CMOS technology.
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号
