The Amdahl ‘s Law is named after computer architect Gene Amdahl , it is used to determine the maximum improvement of a system when only part of it is improved. It states that: The improvement obtained in the performance of a system due to the alteration of one of its components is limited by the fraction of time that component is used. The speed increase of a program using multiple processors in distributed computing is limited by the sequential fraction of the program
Summary
[ hide ]
- 1
- 2
- 3 Amdahl’s Law in parallel teams
- 4 Formula
- 5 Sources
First.
Amdahl’s first law says that the increase in performance due to the inclusion of an improvement with a new resource in the system is limited by the time that improvement is used in the execution of the task.
Second.
Amdahl’s second law says that when an upgrade is introduced to a previously upgraded computer, the performance increase is less than if the upgrade is introduced over the unimproved system. In other words, the incremental improvement in acceleration achieved with the improvement of a part is reduced as new improvements are introduced (Angulo et all, 1995).
Amdahl’s Law on Parallel Teams
Amdahl’s Law is a mathematical model that describes the relationship between the expected acceleration of the parallel implementation of an algorithm and the serial implementation of the same algorithm.
Formula
The original formula of Amdahl’s law is as follows:
F = Fa ((1-Fm) + Fm / Am)
being:
- Fm the improved runtime • Fa the old runtime. • Am the improvement factor that has been introduced in the improved subsystem.
With Amdahl’s formulation the percentage of improvement in the performance of a processor is limited by the percentage of the impact of the element to be modified. Amdahl’s law being generalized is applicable to any problem that has an improvable part and another that cannot be improved.
