Overview of Parallel Architecture CSE Seminar Topic: Parallel computing is a process of computation in which multiple events are being processed and operating on a principle of splitting a large problem into smaller ones which further is processed concurrently. Concurrent indicates parallelism, simultaneity and pipelining.
Parallel events are occurred in many resources during the same time interval; simultaneous events are occurred at the same real time; and pipelined events are occurred in overlapped span of time. Parallel computing deals with concurrent execution of multiple programs in the computer.
The highest level of parallel processing is held among multiple events or programs during multiprogramming, multiprocessing, and time-sharing. Parallel processing is an execution of multiple tasks to solve a computational problem. Most primary reasons for using parallel technology are to save time, executing larger problems, cost savings, transmission speeds, taking advantage of non-local resources etc.
The computational problem has the ability to break a work into discrete pieces of work that is solved simultaneously, to execute multiple tasks at a time, and processed in less time with multiple resources than single resource.
Conclusion:
In today’s world, commercial applications are making use of parallel computers. A computer running an application is able to execute high amount of data in sophisticated method so we can imply that commercial applications may define future parallel computers design. But scientific applications will still remain the users of parallel processing technology. Nowadays, many parallel programming languages and compilers, are splitting a program into multiple processes and/or threads which to be processed concurrently with parallel architectures. Using shared memory or message passing systems, the operating system needs to form communication between the processes and threads. It should also support the operating system that detects, analyzes and manages the dependencies. In order to serialize the concurrent access to the shared resources, mutual exclusion technique is used.
Download Overview of Parallel Architecture CSE Seminar Topic