Journal Article

FZJ-2013-04289

http://join2-wiki.gsi.de/foswiki/pub/Main/Artwork/join2_logo100x88.png Mapping fine-grained power measurements to HPC application runtime characteristics on IBM POWER7

Knobloch, M. (Corresponding author)FZJ* ; Foszczynski, M.FZJ* ; Homberg, W.FZJ* ; Pleiter, D.FZJ* ; Böttiger, H.

2014
Springer Berlin

This record in other databases:

Please use a persistent id in citations: doi:10.1007/s00450-013-0245-5

Abstract: Optimization of energy consumption is a key issue for future HPC. Evaluation of energy consumption requires a fine-grained power measurement. Additional useful information is obtained when performing these measurements at component level. In this paper we describe a setup which allows to perform fine-grained power measurements up to a 1 ms resolution at component level on IBM POWER (IBM and POWER are trademarks of IBM in USA and/or other countries.) machines. We further developed a plug-in for VampirTrace that allows us to correlate these power measurements with application performance characteristics, e.g. obtained by hardware performance counters. This environment enables us to generate both power and performance profiles. Such profiles provide valuable input to develop future strategies for improving workload-driven energy usage per performance. We show in comparison with power profiles of coarser granularity that these fine-grained measurements are necessary to capture the dynamics of power switching.

---

Contributing Institute(s):

1. Jülich Supercomputing Center (JSC)

Research Program(s):

1. 41G - Supercomputer Facility (POF2-41G21) (POF2-41G21)
2. 411 - Computational Science and Mathematical Methods (POF2-411) (POF2-411)
3. ATMLPP - ATML Parallel Performance (ATMLPP) (ATMLPP)

Appears in the scientific report 2014

---