Abstract
Modern high-performance computing platforms, cloud computing systems, and data centers are highly heterogeneous containing nodes where a multicore CPU is tightly integrated with accelerators. An important challenge for energy optimization of hybrid parallel applications on such platforms is how to accurately estimate the energy consumption of application components running on different compute devices of the platform. In this work, we propose a method for accurate estimation of the application component-level energy consumption employing system-level power measurements with power meters. We experimentally validate the method on a cluster of two hybrid heterogeneous computing nodes using three parallel applications - matrix-matrix multiplication, 2D fast Fourier transform and gene sequencing. The experiments demonstrate a high estimation accuracy of the proposed method, with the average estimation error ranging between 2% and 5%. The average error demonstrated by the state-of-the-art estimation methods for the same experimental setup ranges from 15% to 75%, while the maximum reaches 178%. We also show that the use of the state-of-the-art estimation methods instead of the proposed one in the energy optimization loop leads to significant energy losses (up to 45% in our case).
Highlights
High energy consumption by Information and Communications Technology (ICT) systems and devices is a serious concern because of its dire consequences on the economy and environment
We address two challenges for energy optimization of hybrid parallel applications running on modern heterogeneous Non-Uniform Memory Access (NUMA) computing platforms: 1) Accurate modelling of the energy consumptions of application components when executing a hybrid application in parallel on multiple compute devices on a computer
We demonstrate how the dynamic energy consumption can be attributed to the individual application, using AnMoHA, when two different applications are running in parallel on a dual-socket multicore CPU platform
Summary
High energy consumption by Information and Communications Technology (ICT) systems and devices is a serious concern because of its dire consequences on the economy and environment. There exists no solution method to the best of our knowledge that employs system-level power measurements using external power meters to accurately determine the application component level decomposition of the energy consumption of an application executing on multiple independent computing devices in a computer. Using this method, we address two challenges for energy optimization of hybrid parallel applications running on modern heterogeneous NUMA computing platforms: 1) Accurate modelling of the energy consumptions of application components when executing a hybrid application in parallel on multiple compute devices on a computer. A comparison of the accuracy of additive energy models using state-of-the-art on-chip power sensors against the additive energy models constructed with the ground truth for two scientific hybrid applications (matrix-matrix multiplication and 2D fast Fourier transform) on a modern hybrid heterogeneous computing platform containing an Intel multicore CPU, a Nvidia GPU, and an Intel Xeon Phi.
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
