Amdahl’s Law Revisited for Single Chip Systems

Abstract

Amdahl's Law is based upon two assumptions -that of boundlessness and homogeneity -and so it can fail when applied to single chip heterogeneous multiprocessor designs, and even microarchitecture. We show that a performance increase in one part of the system can negatively impact the overall performance of the system, in direct contradiction to the way Amdahl's Law is instructed. Fundamental assumptions that are consistent with Amdahl's Law are a heavily ingrained part of our computing design culture, for research as well as design. This paper points in a new direction. We motivate that emphasis should be made on holistic, system level views instead of divide and conquer approaches. This, in turn, has relevance to the potential impacts of custom processors, system-level scheduling strategies and the way systems are partitioned. We realize that Amdahl's Law is one of the few, fundamental laws of computing. However, its very power is in its simplicity, and if that simplicity is carried over to future systems, we believe that it will impede the potential of future computing systems.

Amdahl's Law in the Multicore Era

“…Furthermore, scheduling software tasks on asymmetric and dynamic multicore chips could be difficult and add overhead. To this end, Tomer Morad and his colleagues 13 and JoAnn Paul and Brett Meyer 14 developed sophisticated models that question the validity of Amdhal's law to future systems, especially embedded ones. On the other hand, more cores might advantageously allow greater parallelism from larger problem sizes, as John Gustafson envisioned.…”

Iso-Energy-Efficiency: An Approach to Power-Constrained Parallel Computation

“…Grama et al [7,10] formally defined isoefficiency as discussed. The fixed-time speedup model [11], memorybounded speedup model [12], and other related studies [13,14] all extend A mdahl's law in unique ways. Ho wever, all of these approaches focus on performance and ignore both energy consumption and the performance effects of power management.…”

“…In (14), and are co mputation and memory access overheads for the ith of p processors in terms of (15) where, nd i l " " r r n u i n f ll workload in all processors.…”

Performance Analysis of Homogeneous On-Chip Large-Scale Parallel Computing Architectures for Data-Parallel Applications

“…In [7], Li and Malek discussed the effect of noncommunication/communication ratio on the speedup based on Amdahl's Law, but his communication delay model is simple without considering the detail of interconnects. In [8], Paul revisited Amdahl's Law on the single chip heterogeneous multiprocessor. His focus was on the performance impact induced by different types of processor cores with different processing capability.…”