On power capping and performance optimization of multithreaded applications

Abstract

Summary Multithreaded applications facilitate the exploitation of the computing power of multicore architectures. On the other hand, these applications can become extremely energy‐intensive, in contrast with the need for limiting the energy usage of computing systems. In this article, we explore the design of techniques enabling multithreaded applications to maximize their performance under a power cap. We consider two control parameters: the number of cores used by the application, and the core power state. We target the design of an autotuning power‐capping technique with minimal intrusiveness and high portability, which is agnostic about the workload profile of the application. We investigate two different approaches for building the strategy for selecting the best configuration of the parameters under control, namely a heuristic approach and a model‐based approach. Through an extensive experimental study, we evaluate the effectiveness of the proposed technique considering two different selection strategies, and we compare them with existing solutions.

DEPO: A dynamic energy‐performance optimizer tool for automatic power capping for energy efficient high‐performance computing

“…In terms of algorithmic comparison of the approach adopted by us to the other solutions described in Section 2 we shall mention that some of the latter differ in adopting an off-line approach, 31 using off-line data, 30 adopt additional code 32 and the majority employ various performance and/or energy models 30,37,38,[42][43][44] for optimization. We propose an on-line approach using auto measurements for runtime tuning without a need for code instrumentation at the cost of a specific application model described above.…”

“…Conoci et al 44 proposed techniques for optimization of throughput of a multithreaded application under a given power cap where two variables are considered: P‐state and the number of threads. The authors tested two strategies: exploration as well as model (with performance and power‐usage model construction) based.…”

Combining Uncore Frequency and Dynamic Power Capping to Improve Power Savings

“…As a consequence, these studies does not rely on dynamic power capping but rather on a static user-defined power cap. Similarly, in [10] the authors adapted the number of cores and the frequency to match a defined power cap. In [33], the authors study the resources configuration to set in order to respect a power cap (hyperthreading, number of cores, .…”