Providing high‐level self‐adaptive abstractions for stream parallelism on multicores

Abstract Stream processing applications are common computing workloads that demand parallelism to increase their performance. As in the past, parallel programming remains a difficult task for application programmers. The complexity increases when programmers must set nonintuitive parameters, is, degree of parallelism. main problem is state‐of‐the‐art libraries use static and not sufficiently abstracted developing stream applications. In this article, we propose self‐adaptive regulation provide higher‐level abstractions. Flexibility provided with two new strategies, one performance experts, other abstracts need goal. We evaluated our solution using compiler transformation rules generate code SPar domain‐specific language. experimental results real‐world highlighted higher abstraction levels without significant degradation comparison executions. strategy experts achieved slightly than works user‐defined goals.