The overall purpose of this paper is to provide an introductory survey in the area of Simultaneous Localization and Mapping (SLAM) particularly its utilization in autonomous vehicle or more specifically in self-driving cars, especially after the release of commercial semiautonomous car like the Tesla vehicles as well as the Google Waymo vehicle. Before we begin diving into the concept of SLAM, we need to understand the importance of SLAM and problems that expand to the various methods developed by numerous researchers to solve it. Thus, in this paper we will start by giving the general concept behind SLAM, followed by sharing details of its different categories and the various methods that form the SLAM function in today's autonomous vehicles; which can solve the SLAM problem. These methods are the current trends that are widely focused in the research community in producing solutions to the SLAM problem; not only in autonomous vehicle but in the robotics field as well. Next, we will compare each of these methods in terms of its pros and cons before concluding the paper by looking at future SLAM challenges.

<span lang="EN-MY">Parallel programming has been implemented in many areas to solve various computational problem with the aim, to improve the performance and scalability of the software application. There are a few parallel programming models commonly used, namely, threads, and message passing (distributed) models. Furthermore, various APIs have been proposed to implement these models based on two popular languages, notably, C/C++ and Java. A few studies have been done to compare the performance of parallel programming models, specifically, pure versus hybrid model. However, most of existing comparisons targeted on MPI/OpenMP based on C/C++ language. In this paper, our aim is to explore the performance comparison between threads, message passing and hybrid model in Java, specifically using Java multithreading and MPJ Express. For this reason, we have chosen a problem called word count occurrence which is significant in Natural Language Processing and use it to design and implement the parallel programs. We then present their performance and discuss the results.</span>