COMPARATIVE ANALYSIS OF MULTIPROCESSOR ARCHITECTURE
Zeeshan Asghar, Usman Ali · International journal of advanced research in computer science and electronics engineering · 2016
Conventionally, the Computer System is viewed to execute the instructions sequentially. From the early days of computers, only certain calculations were made through computers and one by one fetching and execution was made. But, the time has now changed and technology has evolved with time. We are now dealing with huge data centres that perform billions of executions on daily basis, so, sequential execution has no place in the technology world now. In fact, if we go deep into the processor architecture and mechanism, even a sequential machine works parallel, like generating multiple control signals simultaneously. A traditional way to increase performance and speed of a system is to introduce parallel processing that will balance the workload efficiently and will also be responsible for providing reliable services to the user. A Multi-Processor System consist of several processing units sharing multiple memory units and I/O devices to perform high end instructions with reliability. In this paper, we will analyse the Multi-Processor Architectures [1] in detail, that have been a part of a major revolution In the Technology domain and we will explain how one architecture is better in certain situations.