DocumentCode
2786647
Title
Accelerating Vector Based Spatial Analysis by Optimizing Data Transfers
Author
Shulin Cui ; Shuqing Zhang ; Jun Zhang
Author_Institution
Sch. of Adv. Comput., Sun Yat-sen Univ., Guangzhou, China
fYear
2015
fDate
24-26 April 2015
Firstpage
92
Lastpage
95
Abstract
The main goal of this article is to discuss how to accelerate vector based spatial analysis by optimizing data transfers in CUDA C. First we give a new data transfers method based on geometry objects. Then we discuss how to overlap data transfers with computation on GPU. Furthermore we compare the new GPU-based methods with the CPU-based method by using Point-In-Polygon operation. The experimental results show that the new methods are efficient in computation.
Keywords
C language; electronic data interchange; graphics processing units; parallel architectures; CPU-based method; CUDA C; GPU-based method; data transfer method; point-in-polygon operation; vector-based spatial analysis; Acceleration; Computers; Data transfer; Geometry; Graphics processing units; Instruction sets; Kernel; CUDA; GPU-based; overlap data transfers;
fLanguage
English
Publisher
ieee
Conference_Titel
Information Science and Control Engineering (ICISCE), 2015 2nd International Conference on
Conference_Location
Shanghai
Print_ISBN
978-1-4673-6849-0
Type
conf
DOI
10.1109/ICISCE.2015.29
Filename
7120569
Link To Document