국제학회

Distributed Stream Processing System for PTAM on Xeon Phi® Cluster

by 조규남 posted Nov 13, 2014

Prev이전 문서

Next다음 문서

ESC닫기

Extra Form
논문제목	Distributed Stream Processing System for PTAM on Xeon Phi® Cluster
게재일자	2014-11-06
주저자	조규남
공동저자	Chang-sung Jeong
시작페이지,끝페이지	57,64
논문종류	국제 컨퍼런스
국가	대한민국
컨퍼런스명	2014 The International Conference on Future Web (ICFW2014)

Distributed Stream Processing System for PTAM on Xeon Phi® Cluster

Abstract. In this paper, we present a new distributed stream processing system

for PTAM. PTAM has limitation such that its bundle adjustment operation for

map construction cost is extremely high. Our system is designed to overcome

the limitation by deploying distributed stream processing system on cluster

which consists of several servers each with hardware acceleration such as Intel

Xeon Phi. It can help increasing computing power of each server and can

sharply speed up PTAM by cluster.

Keywords: hardware acceleration; GPGPU; Xeon Phi; Distributed stream processing; Parallel computing; Clustering

TAG •

« Prev [SCIE] Cloud-based Mapreduce Workflow Execution Platform

1 2 3

이 사이트를 나눔글꼴로 보기 위해서는
나눔글꼴을 설치해야 합니다.