I developed a parallel array, a static linear data structure for the Standard Template Adaptive Parallel Library (STAPL).
pArray is implemented as a pContainer already defined by the STAPL designers.

My tasks included but were not limited to:

- learning the requirements and constraints of a parallel container;
- coding, debugging, testing and optimising a parallel container;
- determining most needed parallel algorithms;
- coding, debugging, testing and optimizing the algorithms;
- writing a correctness test program to compare the parallel results with sequential algorithms;
- measuring speed up as compared to a sequential data structures;
- comparing efficiency with pVector, a dynamic pContainer;
- writing a report and a presentation.

Here's an abstract for my paper:

pArray as an efficient static parallel container in STAPL

Abstract: Modern computing demands speedy processing of large amounts of data. Using more than one processor at a time greatly increases the speed of program execution, making parallel and distributed processing important. Standard Template Adaptive Parallel Library (STAPL) is a parallel library for C++ that allows users to execute on multiprocessor systems without having to deal with the complexity of application parallelization. STAPL provides distributed data structures with parallel methods, called pContainers (parallel containers). The simplest pContainer STAPL currently provides is a parallel vector, a hard to optimize dynamic data structure. This paper will describe a pArray, a new pContainer that is static and yet sufficient for many applications and provides the highest possible efficiency. pArray builds upon the valarray provided by the Standard Template Library (STL). The basic design will be discussed and performance comparisons with parallel vector will be provided.

contact me