Benefits and Costs of Parallelism: Weighing the Pros and Cons, Exercises of Applications of Computer Sciences

Download Benefits and Costs of Parallelism: Weighing the Pros and Cons and more Exercises Applications of Computer Sciences in PDF only on Docsity! Benefits and costs of parallelism Parallelism is a technique for speeding up a program by running numerous tasks in parallel rather than sequentially, however it comes with "overhead" costs. We must weigh the expenses and rewards of the project and avoid breaking it down into too many little tasks. The extra memory required to accomplish all of the tiny activities at the same time is one example of an added cost. As we all know, parallelism refers to the ability to perform many operations at the same time in order to achieve considerably faster results. It's possible that you'll need to use multi-threading and scatter-gathering to get exceptionally quick results from parallelism. As a result, running speedier programs comes with a variety of costs. In general, there are good, better, and best in all areas. You'll need the core (CPU), the machine (the entire computer), the code (parallelism), the programmer, and potentially the cluster/cloud, for example. logically, parallelism is equivalent to going to Expedia.com and searching for flights, car rentals, and hotels all at the same time. Rather than using a single computer to look for each airline, auto hire, and hotel separately. On the consumer side, having a one-stop shop, so to speak, is more efficient in terms of time and expense. This increases the number of clients who use their services and outweighs the expense of their IT department. References Moseley, T., Shye, A., Reddi, V. J., Grunwald, D., & Peri, R. (2007, March). Shadow profiling: Hiding instrumentation costs with parallelism. In International Symposium on Code Generation and Optimization (CGO'07) (pp. 198-208). IEEE.