Abstract
The memory hierarchy becomes the bottleneck for multiprocessors systems as its evolution does not keep pace with processor technology. This study intends to identify the relationship between performance slow-down and memory pressure, using hardware performance counters. Based on this relationship, we propose an adaptive control system that improves the efficiency of load balancing among the computer resources. The DRAC system, our adaptive control system, observes the access requests on the memory bus. It then adapts its user-level scheduling strategy to maximize the resource utilization. We describe the DRAC system and its mathematical model. We show experimental results that prove the DRAC system is nearly optimal with our model.
Chapter PDF
Similar content being viewed by others
Keywords
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.
References
Bell, R., Malony, A.D., Shende, S.: Paraprof: A portable, extensible, and scalable tool for parallel performance profile analysis. In: Kosch, H., Böszörményi, L., Hellwagner, H. (eds.) Euro-Par 2003. LNCS, vol. 2790, pp. 17–26. Springer, Heidelberg (2003)
Berrendorf, R., Ziegler, H.: Pcl – the performance counter library: A common interface to access hardware performance counters on microprocessors. In Internal Report FZJ-ZAMIB- 9816, Forschungszentrum Julich (1998)
Cappello, F., Richard, O., Etiemble, D.: Investigating the performance of two programming models for clusters of smp pcs. In: Proc. of the 6th Int. Symposium on High Performance Computer Architecture Conference, Toulouse, France, pp. 349–359 (2000)
DeRose, L.: The hardware performance monitor toolkit. In: Sakellariou, R., Keane, J.A., Gurd, J.R., Freeman, L. (eds.) Euro-Par 2001. LNCS, vol. 2150, pp. 122–131. Springer, Heidelberg (2001)
Jin, R., Agrawal, G.: Performance prediction for random write reductions: a case study in modeling shared memory programs. In: Proc. of the 2002 ACM SIGMETRICS int. conf. on Measurement and modeling of computer systems, New York, USA, pp. 117–128 (2002)
London, K., Dongarra, J., Moore, S., Mucci, P., Seymour, K., Spencer, T.: End-user tools for application performance analysis, using hardware counters. In: International Conference on Parallel and Distributed Computing Systems (August 2001)
McCalpin, J.D.: A survey of memory bandwidth and machine balance in current high performance computers. In: IEEE Computer Society Technical Committee on Computer Architecture (TCCA) Newsletter (December 1995)
Miller, B.P.: all. The paradyn parallel performance measurement tool. IEEE Computer 28, 37–46 (1995)
Nikolopoulos, D.S., Polychronopoulos, C.D.: Adaptive scheduling under memory pressure on multiprogrammed clusters. In: Proc. of the Second IEEE/ACM Int. Symp. on Cluster Computing and the Grid (CCGrid 2002), Berlin, Germany (May 2002) (best paper award)
Pattersson, M.: Linux x86 performance-monitoring counters driver, http://user.it.uu.se/~mikpe/linux/perfctr/
Ribler, R.L., Simitci, H., Reed, D.A.: The autopilot performance-directed adaptive control system. Future Generation Computer Systems 18, 175–187 (2001)
SPEC2000. Standard performance evaluation corporation. spec cpu2000 benchmarks (December 1999), http://www.specbench.org/osg/cpu2000
Van-Der-Wijngaart, R.F., Wong, P.: Nas parallel benchmarks i/o version 2.4. In Technical Report NAS-03-002, NASA Ames Research Center (2003)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2004 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Pillon, M., Richard, O., Da Costa, G. (2004). DRAC: Adaptive Control System with Hardware Performance Counters. In: Danelutto, M., Vanneschi, M., Laforenza, D. (eds) Euro-Par 2004 Parallel Processing. Euro-Par 2004. Lecture Notes in Computer Science, vol 3149. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-27866-5_8
Download citation
DOI: https://doi.org/10.1007/978-3-540-27866-5_8
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-22924-7
Online ISBN: 978-3-540-27866-5
eBook Packages: Springer Book Archive