Название: Concurrent, parallel and distributed computingАвтор: Adele Kuzmiakova
Издательство: Arcler Press
Год: 2023
Страниц: 260
Язык: английский
Формат: pdf (true)
Размер: 10.1 MB
The book "Concurrent, Parallel, and Distributed Computing" offers an excellent overview of the various areas of the computing field. There is a lot of overlap between the words "concurrent computing," "parallel computing," and "distributed computing," and there is no obvious differentiation between them. The same system can be described as "parallel" and "distributed"; in a typical distributed system, the processors run concurrently in parallel.The content in the book is presented in such a way that even a reader with no prior knowledge of computers may understand it and become acquainted with the fundamental concepts of computing. It offers numerous small examples, demonstration materials, and sample exercises that teachers can use to teach parallel programming principles to students who have just recently been introduced to basic programming concepts. It focuses on Python multiprocessing features like fork/join threading, message passing, sharing resources between threads, and using locks. Parallelism's utility can be seen in applications like searching, sorting, and simulations. Students and researchers can get an accessible and comprehensive explanation of the concepts, guidelines, and, in particular, the complex instrumentation techniques used in computing.
However, the concurrency is inherently difficult because it is not very straightforward to identify racing circumstances via experimentation. Even if a test finds a flaw, pinpointing the section of the software that is producing it might be difficult. Concurrency difficulties are notoriously hard to replicate. It is tough to get them to occur in the same method recurrently. The comparative accuracy of predictions that are highly impacted by the atmosphere determines how commands or information are interleaved. Additional running applications, other network activity, system software development selections, differences in computer timepiece speed, and so on may all generate stays. You could obtain different results each time you execute a program with compatibility issues. These are heisenbugs, which are difficult to replicate, as contrasted to “bohrbugs,” which appear every time you look at them. Bohrbugs account for almost all bugs in linear development. Concurrency is difficult to master. However, do not let this deter you. We will examine rational techniques to build concurrent software that is secure from such types of issues during the following many readings.
Скачать Concurrent, parallel and distributed computing