Utilizing Heterogeneous Memory Hierarchies in the PGAS Model
Authors:
- Roger KowalewskiRKRoger Kowalewski
- Tobias FuchsTFTobias Fuchs
- Karl FurlingerKFKarl Furlinger
- Tobias GuggemosTGTobias Guggemos
- 1 March 2018
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 353-357
Abstract
Emerging technologies such as non-volatile or 3D-stacked memory significantly impact the design of future high performance computing systems. To keep up with the increasing core count, relying only on DRAM is inefficient due to its static power consumption. Modern HPC architectures feature a heterogeneous memory hierarchy with different capacities and capabilities. This paper addresses two challenges. First, we need programming models to abstract the complexity of the underlying heterogeneous memory hierarchy, while still giving explicit control to domain experts. We propose the concept of memory spaces to model a heterogeneous memory hierarchy and integrate it into a PGAS-like programming model. Second, we need to understand the impact of different memory capabilities on the performance of scientific applications. An experimental evaluation with a series of benchmarks, conducted on a Intel KNL platform, reveals that proper data placement on specific types of memory achieves significant speedup.Keywords
This publication has 12 references indexed in Scilit:
- DASH: A C++ PGAS Library for Distributed Data Structures and Parallel Algorithms2016
- HpMC2015
- Heterogeneous memory architectures: A HW/SW approach for mixing die-stacked and off-package memories2015
- CAMEO: A Two-Level Memory Organization with Capacity of Main Memory and Flexibility of Hardware-Managed Cache2014
- Benchmarking Usability and Performance of Multicore Languages2013
- Hybrid memory cube new DRAM architecture increases density and performance2012
- Introducing OpenSHMEM2010
- Architecting phase change memory as a scalable dram alternativeACM SIGARCH Computer Architecture News, 2009
- Parallel Programmability and the Chapel LanguageThe International Journal of High Performance Computing Applications, 2007
- X10ACM SIGPLAN Notices, 2005