Power use of disk subsystems in supercomputers

Exascale will present many challenges to the HPC community, but the primary problem will likely be power consumption. Current petascale systems already use a significant fraction of the power that an exascale system will be allotted. In this paper, we show measurements for real I/O power use in three large systems. We show that I/O power use is proportionally fairly low per machine, between 4.4 and 5.5% of the total consumption. We use these measurements to motivate a burst-buffer checkpointing solution for power-efficient I/O at exascale. We estimated this solution to use approximately 6.6% of the exascale machine power budget, which is on par with today's systems.

[1]  James A. Ang,et al.  The Alliance for Computing at the Extreme Scale. , 2010 .

[2]  Bruce Jacob,et al.  Memory Systems: Cache, DRAM, Disk , 2007 .

[3]  David F. Heidel,et al.  An Overview of the BlueGene/L Supercomputer , 2002, ACM/IEEE SC 2002 Conference (SC'02).

[4]  Hsing-bung Chen,et al.  An Intelligent Parallel and Scalable Server I/O Networking Environment for High Performance Cluster Computing Systems , 2008, PDPTA.

[5]  Hyosung Kim,et al.  300V DC feed system for Internet data center , 2011, 8th International Conference on Power Electronics - ECCE Asia.

[6]  Chita R. Das,et al.  The Penn State Computing Condominium Scheduling System , 1998, Proceedings of the IEEE/ACM SC98 Conference.

[7]  Euiseong Seo,et al.  Empirical Analysis on Energy Efficiency of Flash-based SSDs , 2008, HotPower.

[8]  Jack J. Dongarra,et al.  The LINPACK Benchmark: past, present and future , 2003, Concurr. Comput. Pract. Exp..