A Survey on Amdahl's Law Extension in Multicore Architectures

Multicore architectures represent the future of computing since they provide cost effective solutions to improve throughput and performance of parallel programs while keeping power consumption manageable. Amdahl’s law continues to serve as a guideline for parallel programmers to assess the upper bounds of attainable performance in multicore architectures. In this article, we review the key papers related to the extension of Amdahl’s law for multicore architectures by characterizing them into five categories. For each category we briefly survey the main analytic modeling techniques and discuss their inherent advantages and disadvantages. All the analytic models discussed in this article are compared against a number of attributes hindering the performance enhancement of multicore computing to pinpoint their strengths and weaknesses. Finally, we recommend directions for future work to inspire the research community to invent new ideas to model and evaluate the emerging multicore computing paradigms.