SIA OpenIR  > 数字工厂研究室
A Synchronization Mechanism between CUDA Blocks for GPU
Wang, Bingru; Zhang, Changyou; Wang F(王锋); Feng, Jun
作者部门数字工厂研究室
会议名称2nd International Conference on Control, Automation and Artificial Intelligence (CAAI)
会议日期June 25-26, 2017
会议地点Sanya, CHINA
会议主办者Sci & Engn Res Ctr
会议录名称PROCEEDINGS OF THE 2017 2ND INTERNATIONAL CONFERENCE ON CONTROL, AUTOMATION AND ARTIFICIAL INTELLIGENCE (CAAI 2017)
出版者ATLANTIS PRESS
出版地PARIS
2017
页码251-254
收录类别CPCI(ISTP)
WOS记录号WOS:000426676200056
产权排序3
ISSN号1951-6851
关键词Gpu Synchronization Mechanism Sssp Parallel Computing Delta-stepping Cuda
摘要GPU(Graphic Processing Unit) provides a promising solution with massive threads and its advantage is high performance computing. The emergence of CUDA(Compute Unified Device Architecture) opens the door of using GPU's powerful computing power. However, because of the limitation of CUDA itself, direct communication is not supported between SMs(streaming multiprocessors) on GPU. It is time-consuming by atomic operation or barrier synchronization. A synchronization mechanism has been proposed in this paper, that is, on the premise of result available, the times of kernel launched should be reduced. Each kernel launched, it should be computed enough on GPU, the results back to the CPU. Based on SSSP, the validity of this method is illustrated by delta-stepping. For facebook dataset, compared with atomic operation, the speedup ratio is about 1.8. For New York map dataset, compared with atomic operation and barrier synchronization, the speedup ratio is about 9.3 and 1.7 separately.
语种英语
引用统计
文献类型会议论文
条目标识符http://ir.sia.cn/handle/173321/21555
专题数字工厂研究室
通讯作者Feng, Jun
作者单位1.Shijiazhuang Tiedao University, Shijiazhuang, China
2.Institute of Software, Chinese Academy of Science, Beijing, China
3.Shenyang Institute of Automation, Chinese Academy of Science, Shenyang, China
推荐引用方式
GB/T 7714
Wang, Bingru,Zhang, Changyou,Wang F,et al. A Synchronization Mechanism between CUDA Blocks for GPU[C]//Sci & Engn Res Ctr. PARIS:ATLANTIS PRESS,2017:251-254.
条目包含的文件 下载所有文件
文件名称/大小 文献类型 版本类型 开放类型 使用许可
A Synchronization Me(2107KB)会议论文 开放获取CC BY-NC-SA浏览 下载
个性服务
推荐该条目
保存到收藏夹
查看访问统计
导出为Endnote文件
谷歌学术
谷歌学术中相似的文章
[Wang, Bingru]的文章
[Zhang, Changyou]的文章
[Wang F(王锋)]的文章
百度学术
百度学术中相似的文章
[Wang, Bingru]的文章
[Zhang, Changyou]的文章
[Wang F(王锋)]的文章
必应学术
必应学术中相似的文章
[Wang, Bingru]的文章
[Zhang, Changyou]的文章
[Wang F(王锋)]的文章
相关权益政策
暂无数据
收藏/分享
文件名: A Synchronization Mechanism between CUDA Blocks for GPU.pdf
格式: Adobe PDF
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。