Abstract
We consider the problem of energy-optimally mapping a set of moldable-parallel tasks in the steady-state pattern of a software-pipelined streaming computation onto a generic many-core CPU architecture with a 2D mesh geometry, where the execution voltage and frequency levels of the cores can be selected dynamically from a given set of discrete DVFS levels. We extend the Crown Scheduling technique for parallelizable tasks to temperature-aware scheduling, taking into account the tasks’ heat generation, the heat limit for each core, and the heat diffusion along the 2D mesh geometry of typical many-core CPU architectures. Our approach introduces a systematic method for alternating task executions between disjoint “buddy” core groups in subsequent iterations of crown schedules to avoid long-time overheating of cores. We present two integer linear program (ILP) solutions with different degrees of flexibility, and show that these can be solved for realistic problem sizes with today’s ILP solver technology. Experiments with several streaming task graphs derived from real-world applications show that the flexibility for the scheduler can be greatly increased by considering buddy-cores, thus finding feasible solutions in scenarios that could not be solved otherwise. We also present a fast heuristic for the same problem.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
We assume that tasks are computation-bound, i.e. that task runtime is inverse to core frequency, so that decisions on resource allocation and frequency scaling can be separated. Extensions to memory-bound or communication-bound tasks are possible.
- 2.
Core allocations must be multiples of 2, which is automatically preserved by crown scheduling if \(w_{j,v}\ge 2\).
- 3.
In (rather unlikely) borderline cases it could actually make sense to run even a cold task in alternating mode to make it even colder, so that it can compensate for the heat impact of a hot task on the same core(s) without having to use the more inefficient alternating variant for that hot task.
- 4.
If the penalty might differ between tasks, then the misscost table could be further indexed by the task index.
References
Alkabani, Y., Koushanfar, F., Potkonjak, M.: N-version temperature-aware scheduling and binding. In: Proceedings of the International Symposium on Low Power Electronics and Design, San Francisco, CA, USA, pp. 331–334. ACM, August 2009
Bampis, E., Letsios, D., Lucarelli, G., Markakis, E., Milis, I.: On multiprocessor temperature-aware scheduling problems. In: Snoeyink, J., Lu, P., Su, K., Wang, L. (eds.) AAIM/FAW -2012. LNCS, vol. 7285, pp. 149–160. Springer, Heidelberg (2012). https://doi.org/10.1007/978-3-642-29700-7_14
Bao, M., Andrei, A., Eles, P.I., Peng, Z.: Temperature-aware task mapping for energy optimization with dynamic voltage scaling. In: 11th IEEE Workshop on Design and Diagnostics of Electronic Circuits and Systems, pp. 44–49 (2008)
Bao, M., Andrei, A., Eles, P.I., Peng, Z.: Temperature-aware idle time distribution for energy optimization with dynamic voltage scaling. In: Proceedings of the Design, Automation, and Test in Europe Conference and Exhibition (DATE 2010), pp. 21–26 (2010)
Chantem, T., Hu, X.S., Dick, R.P.: Temperature-aware scheduling and assignment for hard real-time applications on MPSoCs. IEEE Trans. Very Large Scale Integr. Syst. 19(10), 1884–1897 (2011). https://doi.org/10.1109/TVLSI.2010.2058873
Coskun, A.K., Rosing, T.S., Whisnant, K.: Temperature aware task scheduling in MPSoCs. In: Proceedings of the DATE 2007 (2007)
Eindhoven Technical University, Electronic Systems: Dataflow Benchmark Suite (DFbench) (2010). http://www.es.ele.tue.nl/dfbench/
Eitschberger, P.: Energy-efficient and fault-tolerant scheduling for manycores and grids. Ph.D. thesis, FernUniversität in Hagen, Germany (2017)
Holmbacka, S., Keller, J.: Workload type-aware scheduling on big.LITTLE platforms. In: Ibrahim, S., Choo, K.-K.R., Yan, Z., Pedrycz, W. (eds.) ICA3PP 2017. LNCS, vol. 10393, pp. 3–17. Springer, Cham (2017). https://doi.org/10.1007/978-3-319-65482-9_1
Hällis, F., Holmbacka, S., Lund, W., Slotte, R., Lafond, S., Lilius, J.: Thermal influence on the energy efficiency of workload consolidation in many-core architectures. In: 24th Tyrrhenian International Workshop Digital Communications-Green ICT, pp. 1–6 (2013)
Jaja, J.: An Introduction to Parallel Algorithms. Addison Wesley, Boston (1992)
Jayaseelan, R., Mitra, T.: Temperature aware scheduling for embedded processors. J. Low Power Electron. 5(3), 363–372 (2009)
Kahn, G.: The semantics of a simple language for parallel programming. In: Proceedings of the IFIP Congress on Information Processing, pp. 471–475. North-Holland (1974)
Kessler, C., Litzinger, S., Keller, J.: Static scheduling of moldable streaming tasks with task fusion for parallel systems with DVFS. IEEE Trans. Comput.-Aided Des. Integr. Circuits Syst. (TCAD) 39(11), 4166–4178 (2020)
Litzinger, S., Keller, J.: Influence of discretization of frequencies and processor allocation on static scheduling of parallelizable tasks with deadlines. PARS-Mitteilungen 35(1), 95–108 (2020)
Lu, S.J., Tessier, R., Burleso, W.: Reinforcement learning for thermal-aware many-core task allocation. In: Proceedings of the GLSVLSI 2015. ACM, May 2015
Melot, N., Kessler, C., Eitschberger, P., Keller, J.: Co-optimizing core allocation, mapping and DVFS in streaming programs with moldable tasks for energy efficient execution on manycore architectures. In: Proceedings of the 19th International Conference on Application of Concurrency to System Design (ACSD 2019) (2019)
Melot, N., Kessler, C., Keller, J., Eitschberger, P.: Fast crown scheduling heuristics for energy-efficient mapping and scaling of moldable streaming tasks on manycore systems. ACM Trans. Archit. Code Optim. 11(4), 1–24 (2015)
Pierson, J., Stolf, P., Sun, H., Casanova, H.: MILP formulations for spatio-temporal thermal-aware scheduling in cloud and HPC datacenters. Clust. Comput. 23(2), 421–439 (2020). https://doi.org/10.1007/s10586-019-02931-3
Rajan, D., Yu, P.S.: Temperature-aware scheduling: When is system-throttling good enough? In: Proceedings of the 9th International Conference on Web-Age Information Management (WAIM 2008), Zhangjiajie, China, pp. 397–404. IEEE Computer Society, July 2008
Acknowledgments
C. Kessler acknowledges partial funding by ELLIIT, project GPAI.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 Springer Nature Switzerland AG
About this paper
Cite this paper
Kessler, C., Keller, J., Litzinger, S. (2021). Temperature-Aware Energy-Optimal Scheduling of Moldable Streaming Tasks onto 2D-Mesh-Based Many-Core CPUs with DVFS. In: Klusáček, D., Cirne, W., Rodrigo, G.P. (eds) Job Scheduling Strategies for Parallel Processing. JSSPP 2021. Lecture Notes in Computer Science(), vol 12985. Springer, Cham. https://doi.org/10.1007/978-3-030-88224-2_9
Download citation
DOI: https://doi.org/10.1007/978-3-030-88224-2_9
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-88223-5
Online ISBN: 978-3-030-88224-2
eBook Packages: Computer ScienceComputer Science (R0)