Date of Publication :17th January 2024

Abstract: Although hardware manufacturers are relentlessly building multi-cores system, software parallelization will not advance simply because the hardware has been built that way; it is a conscious effort of developing frameworks and models that can optimally exploit inherent hardware parallelism and make software implementation easy. Thus, parallel programming is an essential concern of software engineers in order to meet up with the call -in technological development. In the generation where there is incredibly massive bulk data to deal with, software developers need to seek algorithms or techniques that will directly take advantage of the underlying parallelized processing resources. This work focuses on the design of a fork and join parallel programming framework which can handle bulk data operations inherent in enterprise solutions. Techniques for designing parallel framework was followed and Amdahl’s law is adopted to identify the appreciable speedup point so as to know the quantity of threads that could give optimum performance at any given time while exploiting available processor cores.

Reference :