Student:	Jalil Ahmad FAZELY
Supervisor:	Asst. Prof. Dr. Ibrahim SAVRAN
Department:	Bilgisayar Mühendisliği
Institution:	Graduate School of Natural and Applied Sciences
University:	Karadeniz Technical University Turkey

Title of the Thesis:	HIGH PERFORMANCE COMPUTING ON BIG DATA IN GEOGRAPHIC
Level:	M.Sc.
Acceptance Date:	18/6/2019
Number of Pages:	29
Registration Number:	i3611

Summary:

Over recent decades, High-performance computing helps researchers to manipulate

big data. Thus, in this paper, we proposed and implemented two methods to produce

Delaunay triangulation and the convex hull on big data in Geographic Information Systems

(GIS). Our methods are dividing the set of spatial points into non-collapsing blocks. Each

block is triangulated separately. Finally, in order to concatenate these blocks, we used our

sewing methods. For a set of twenty million points on the 2D surface, we achieved 1.92x

speedup over the traditional method.

Keywords: Delaunay Triangulation; Geographic Information System (GIS); Big Data; 2D;

Sewing-Method; Convex-Hull; Merging; High Performance Computing.