Thread-Count Tuning for Carbon-Efficient Multicore HPC Applications
Resumo
As multicore HPC systems expose increasingly large degrees of parallelism, using more hardware resources does not necessarily translate into more useful scientific computation. Once application scalability saturates, activating additional threads can provide little or no performance benefit while increasing energy consumption and the carbon cost of computation. Therefore, understanding this trade-off is essential for improving resource efficiency and advancing sustainable HPC. In this paper, we quantify how thread-count tuning affects performance and operational carbon emissions across 20 OpenMP applications and five multicore CPU platforms spanning 16–128 hardware threads. A full design-space exploration identifies performance-, carbon-, and joint-optimal configurations. Relative to maximum-thread execution, performance-optimal tuning improves geometric-mean runtime by up to 16.0%, while carbon-optimal tuning reduces emissions by up to 27.8%. Scalability-limited workloads benefit substantially more: sp.C reduces emissions by 69.5% on AMD128.Referências
Anthony, L. F. W., Kanding, B., and Selvan, R. (2020). Carbontracker: Tracking and predicting the carbon footprint of training deep learning models. ICML Workshop on Challenges in Deploying and monitoring Machine Learning Systems. arXiv:2007.03051.
Azhar, M. W., Manivannan, M., and Stenström, P. (2023). Approx-rm: Reducing energy on heterogeneous multicore processors under accuracy and timing constraints. ACM Trans. Archit. Code Optim., 20(3).
Bailey, D., Harris, T., Saphir, W., Van Der Wijngaart, R., Woo, A., and Yarrow, M. (1995). The nas parallel benchmarks 2.0. Technical report, Technical Report NAS-95-020, NASA Ames Research Center.
Brigham, E. O. and Morrow, R. E. (1967). The fast fourier transform. IEEE Spectrum, 4(12):63–70.
Cabrera, A., Almeida, F., Castellanos-Nieves, D., Oleksiak, A., and Blanco, V. (2023). Energy efficient power cap configurations through pareto front analysis and machine learning categorization. Cluster Computing, 27(3):3433–3449.
Chadha, G., Mahlke, S., and Narayanasamy, S. (2012). When less is more (LIMO): Controlled parallelism for improved efficiency. In CASES, pages 141–150, USA.
Che, S., Boyer, M., Meng, J., Tarjan, D., Sheaffer, J. W., Lee, S.-H., and Skadron, K. (2009). Rodinia: A benchmark suite for heterogeneous computing. In 2009 IEEE International Symposium on Workload Characterization (IISWC), pages 44–54.
Chheda, S., Verma, G., Tian, S., Chapman, B., and Doerfert, J. (2024). Evaluating tuning opportunities of the llvm/openmp runtime. In SC24-W: Workshops of the International Conference for High Performance Computing, Networking, Storage and Analysis, pages 919–929.
Climate, E. (2024). Global electricity review 2024. [link]. Accessed: 2024-04-01.
EIA (2023). State electricity profiles - tennessee. [link]. U.S. Energy Information Administration. Accessed: 2024-04-01.
Etinski, M., Corbalan, J., Labarta, J., and Valero, M. (2010). Optimizing job performance under a given power constraint in hpc centers. In International Conference on Green Computing, pages 257–267.
Gupta, U., Elgamal, M., Hills, G., Wei, G.-Y., Lee, H.-H. S., Brooks, D., and Wu, C.-J. (2022). Act: designing sustainable computer systems with an architectural carbon modeling tool. In Proceedings of the 49th Annual International Symposium on Computer Architecture, ISCA ’22, page 784–799, New York, NY, USA. Association for Computing Machinery.
Heroux, M. A., Dongarra, J., and Luszczek, P. (2013). Hpcg benchmark technical specification.
IEA (2023). Electricity information 2023: Overview. [link]. International Energy Agency. Accessed: 2024-04-01.
IESO (2023). Ontario’s electricity data. [link]. Independent Electricity System Operator. Accessed: 2024-04-01.
Karlin, I., Bhatele, A., Keasler, J., Chamberlain, B. L., Cohen, J., DeVito, Z., Haque, R., Laney, D., Luke, E., Wang, F., Richards, D., Schulz, M., and Still, C. (2013). Exploring traditional and emerging parallel programming models using a proxy application. In 27th IEEE International Parallel & Distributed Processing Symposium (IEEE IPDPS 2013), Boston, USA.
Kline, D., Parshook, N., Ge, X., Brunvand, E., Melhem, R., Chrysanthis, P. K., and Jones, A. K. (2019). Greenchip: A tool for evaluating holistic sustainability of modern computing systems. Sustainable Computing: Informatics and Systems, 22:322–332.
Krzywaniak, A., Czarnul, P., and Proficz, J. (2022). Depo: A dynamic energy-performance optimizer tool for automatic power capping for energy efficient high-performance computing. Software: Practice and Experience, 52(12):2598–2634.
Lannelongue, L., Grealey, J., and Inouye, M. (2021). Green algorithms: Quantifying the carbon footprint of computation. Advanced Science, 8.
Li, B., Basu Roy, R., Wang, D., Samsi, S., Gadepally, V., and Tiwari, D. (2023). Toward sustainable hpc: Carbon footprint estimation and environmental implications of hpc systems. In Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis, SC ’23, New York, NY, USA. Association for Computing Machinery.
Lorenzon, A. F., De Oliveira, C. C., Souza, J. D., and Beck, A. C. S. (2018). Aurora: Seamless optimization of openmp applications. IEEE transactions on parallel and distributed systems, 30(5):1007–1021.
Lorenzon, A. F., Souza, J. D., and Beck, A. C. S. (2017). Laant: A library to automatically optimize edp for openmp applications. In DATE, pages 1229–1232. IEEE.
Maji, D., Pfaff, B., PR, V., Sreenivasan, R., Firoiu, V., Iyer, S., Josephson, C., Pan, Z., and Sitaraman, R. K. (2023). Bringing carbon awareness to multi-cloud application delivery. In Proceedings of the 2nd Workshop on Sustainable Computer Systems, pages 1–6.
Marathe, A., Bailey, P. E., Lowenthal, D. K., Rountree, B., Schulz, M., and de Supinski, B. R. (2015). A run-time system for power-constrained hpc applications. In Kunkel, J. M. and Ludwig, T., editors, High Performance Computing, pages 394–408, Cham. Springer.
National Grid ESO (2023). Carbon intensity api and national grid dashboard. [link]. National Grid Electricity System Operator. Accessed: 2024-04-01.
Navarro Muñoz, A., F. Lorenzon, A., Ayguadé Parra, E., and Beltran Querol, V. (2021). Combining dynamic concurrency throttling with voltage and frequency scaling on task-based programming models. In Proceedings of the 50th International Conference on Parallel Processing, ICPP ’21, New York, NY, USA. Association for Computing Machinery.
ONS (2023). Boletim mensal de geração - brasil. [link]. Operador Nacional do Sistema Elétrico. Accessed: 2024-04-01.
Raasch, S. E. and Reinhardt, S. K. (2003). The impact of resource partitioning on smt processors. In PACT, pages 15–25.
RTE (2023). Bilan Électrique 2023. [link]. Réseau de Transport d’Électricité. Accessed: 2024-04-01.
Sensi, D. D., Torquati, M., and Danelutto, M. (2016). A reconfiguration algorithm for power-aware parallel applications. TACO, 13(4):43:1–43:25.
Silveira, A. and Lorenzon, A. (2024). Análise da pegada de carbono em aplicações paralelas. In Anais da XXIV Escola Regional de Alto Desempenho da Região Sul, pages 111–112, Porto Alegre, RS, Brasil. SBC.
Sridharan, S., Gupta, G., and Sohi, G. S. (2014). Adaptive, efficient, parallel execution of parallel programs. In ACM SIGPLAN PLDI, pages 169–180, NY, USA. ACM.
Subramanian, L., Seshadri, V., Kim, Y., Jaiyen, B., and Mutlu, O. (2013). Mise: Providing performance predictability and improving fairness in shared main memory systems. In IEEE HPCA, pages 639–650.
Suleman, M. A., Qureshi, M. K., and Patt, Y. N. (2008). Feedback-driven threading: power-efficient and high-performance execution of multi-threaded workloads on cmps. SIGPLAN Not., 43(3):277–286.
Terna (2023). Dati statistici sull’energia elettrica in italia. [link]. Terna - Rete Elettrica Nazionale. Accessed: 2024-04-01.
Tomorrow (2024). electricitymap - live CO2 emissions of electricity consumption. [link]. Accessed: 2024-04-01.
Wang, C., Zink, M., and Irwin, D. (2018). Energy-agile design for parallel hpc applications. Sustainable Computing: Informatics and Systems, 19:123–134.
Zhang, H. and Hoffmann, H. (2016). Maximizing performance under a power cap: A comparison of hardware, software, and hybrid techniques. SIGARCH Comput. Archit. News, 44(2):545–559.
Zhou, Z., Liu, F., Xu, Y., Zou, R., Xu, H., Lui, J. C., and Jin, H. (2013). Carbon-aware load balancing for geo-distributed cloud services. In 2013 IEEE 21st International Symposium on Modelling, Analysis and Simulation of Computer and Telecommunication Systems, pages 232–241.
Azhar, M. W., Manivannan, M., and Stenström, P. (2023). Approx-rm: Reducing energy on heterogeneous multicore processors under accuracy and timing constraints. ACM Trans. Archit. Code Optim., 20(3).
Bailey, D., Harris, T., Saphir, W., Van Der Wijngaart, R., Woo, A., and Yarrow, M. (1995). The nas parallel benchmarks 2.0. Technical report, Technical Report NAS-95-020, NASA Ames Research Center.
Brigham, E. O. and Morrow, R. E. (1967). The fast fourier transform. IEEE Spectrum, 4(12):63–70.
Cabrera, A., Almeida, F., Castellanos-Nieves, D., Oleksiak, A., and Blanco, V. (2023). Energy efficient power cap configurations through pareto front analysis and machine learning categorization. Cluster Computing, 27(3):3433–3449.
Chadha, G., Mahlke, S., and Narayanasamy, S. (2012). When less is more (LIMO): Controlled parallelism for improved efficiency. In CASES, pages 141–150, USA.
Che, S., Boyer, M., Meng, J., Tarjan, D., Sheaffer, J. W., Lee, S.-H., and Skadron, K. (2009). Rodinia: A benchmark suite for heterogeneous computing. In 2009 IEEE International Symposium on Workload Characterization (IISWC), pages 44–54.
Chheda, S., Verma, G., Tian, S., Chapman, B., and Doerfert, J. (2024). Evaluating tuning opportunities of the llvm/openmp runtime. In SC24-W: Workshops of the International Conference for High Performance Computing, Networking, Storage and Analysis, pages 919–929.
Climate, E. (2024). Global electricity review 2024. [link]. Accessed: 2024-04-01.
EIA (2023). State electricity profiles - tennessee. [link]. U.S. Energy Information Administration. Accessed: 2024-04-01.
Etinski, M., Corbalan, J., Labarta, J., and Valero, M. (2010). Optimizing job performance under a given power constraint in hpc centers. In International Conference on Green Computing, pages 257–267.
Gupta, U., Elgamal, M., Hills, G., Wei, G.-Y., Lee, H.-H. S., Brooks, D., and Wu, C.-J. (2022). Act: designing sustainable computer systems with an architectural carbon modeling tool. In Proceedings of the 49th Annual International Symposium on Computer Architecture, ISCA ’22, page 784–799, New York, NY, USA. Association for Computing Machinery.
Heroux, M. A., Dongarra, J., and Luszczek, P. (2013). Hpcg benchmark technical specification.
IEA (2023). Electricity information 2023: Overview. [link]. International Energy Agency. Accessed: 2024-04-01.
IESO (2023). Ontario’s electricity data. [link]. Independent Electricity System Operator. Accessed: 2024-04-01.
Karlin, I., Bhatele, A., Keasler, J., Chamberlain, B. L., Cohen, J., DeVito, Z., Haque, R., Laney, D., Luke, E., Wang, F., Richards, D., Schulz, M., and Still, C. (2013). Exploring traditional and emerging parallel programming models using a proxy application. In 27th IEEE International Parallel & Distributed Processing Symposium (IEEE IPDPS 2013), Boston, USA.
Kline, D., Parshook, N., Ge, X., Brunvand, E., Melhem, R., Chrysanthis, P. K., and Jones, A. K. (2019). Greenchip: A tool for evaluating holistic sustainability of modern computing systems. Sustainable Computing: Informatics and Systems, 22:322–332.
Krzywaniak, A., Czarnul, P., and Proficz, J. (2022). Depo: A dynamic energy-performance optimizer tool for automatic power capping for energy efficient high-performance computing. Software: Practice and Experience, 52(12):2598–2634.
Lannelongue, L., Grealey, J., and Inouye, M. (2021). Green algorithms: Quantifying the carbon footprint of computation. Advanced Science, 8.
Li, B., Basu Roy, R., Wang, D., Samsi, S., Gadepally, V., and Tiwari, D. (2023). Toward sustainable hpc: Carbon footprint estimation and environmental implications of hpc systems. In Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis, SC ’23, New York, NY, USA. Association for Computing Machinery.
Lorenzon, A. F., De Oliveira, C. C., Souza, J. D., and Beck, A. C. S. (2018). Aurora: Seamless optimization of openmp applications. IEEE transactions on parallel and distributed systems, 30(5):1007–1021.
Lorenzon, A. F., Souza, J. D., and Beck, A. C. S. (2017). Laant: A library to automatically optimize edp for openmp applications. In DATE, pages 1229–1232. IEEE.
Maji, D., Pfaff, B., PR, V., Sreenivasan, R., Firoiu, V., Iyer, S., Josephson, C., Pan, Z., and Sitaraman, R. K. (2023). Bringing carbon awareness to multi-cloud application delivery. In Proceedings of the 2nd Workshop on Sustainable Computer Systems, pages 1–6.
Marathe, A., Bailey, P. E., Lowenthal, D. K., Rountree, B., Schulz, M., and de Supinski, B. R. (2015). A run-time system for power-constrained hpc applications. In Kunkel, J. M. and Ludwig, T., editors, High Performance Computing, pages 394–408, Cham. Springer.
National Grid ESO (2023). Carbon intensity api and national grid dashboard. [link]. National Grid Electricity System Operator. Accessed: 2024-04-01.
Navarro Muñoz, A., F. Lorenzon, A., Ayguadé Parra, E., and Beltran Querol, V. (2021). Combining dynamic concurrency throttling with voltage and frequency scaling on task-based programming models. In Proceedings of the 50th International Conference on Parallel Processing, ICPP ’21, New York, NY, USA. Association for Computing Machinery.
ONS (2023). Boletim mensal de geração - brasil. [link]. Operador Nacional do Sistema Elétrico. Accessed: 2024-04-01.
Raasch, S. E. and Reinhardt, S. K. (2003). The impact of resource partitioning on smt processors. In PACT, pages 15–25.
RTE (2023). Bilan Électrique 2023. [link]. Réseau de Transport d’Électricité. Accessed: 2024-04-01.
Sensi, D. D., Torquati, M., and Danelutto, M. (2016). A reconfiguration algorithm for power-aware parallel applications. TACO, 13(4):43:1–43:25.
Silveira, A. and Lorenzon, A. (2024). Análise da pegada de carbono em aplicações paralelas. In Anais da XXIV Escola Regional de Alto Desempenho da Região Sul, pages 111–112, Porto Alegre, RS, Brasil. SBC.
Sridharan, S., Gupta, G., and Sohi, G. S. (2014). Adaptive, efficient, parallel execution of parallel programs. In ACM SIGPLAN PLDI, pages 169–180, NY, USA. ACM.
Subramanian, L., Seshadri, V., Kim, Y., Jaiyen, B., and Mutlu, O. (2013). Mise: Providing performance predictability and improving fairness in shared main memory systems. In IEEE HPCA, pages 639–650.
Suleman, M. A., Qureshi, M. K., and Patt, Y. N. (2008). Feedback-driven threading: power-efficient and high-performance execution of multi-threaded workloads on cmps. SIGPLAN Not., 43(3):277–286.
Terna (2023). Dati statistici sull’energia elettrica in italia. [link]. Terna - Rete Elettrica Nazionale. Accessed: 2024-04-01.
Tomorrow (2024). electricitymap - live CO2 emissions of electricity consumption. [link]. Accessed: 2024-04-01.
Wang, C., Zink, M., and Irwin, D. (2018). Energy-agile design for parallel hpc applications. Sustainable Computing: Informatics and Systems, 19:123–134.
Zhang, H. and Hoffmann, H. (2016). Maximizing performance under a power cap: A comparison of hardware, software, and hybrid techniques. SIGARCH Comput. Archit. News, 44(2):545–559.
Zhou, Z., Liu, F., Xu, Y., Zou, R., Xu, H., Lui, J. C., and Jin, H. (2013). Carbon-aware load balancing for geo-distributed cloud services. In 2013 IEEE 21st International Symposium on Modelling, Analysis and Simulation of Computer and Telecommunication Systems, pages 232–241.
Publicado
03/11/2026
Como Citar
SILVEIRA, Arthur; COSTA, Matheus; OSTHOFF, Carla; NAVAUX, Philippe O. A.; LORENZON, Arthur.
Thread-Count Tuning for Carbon-Efficient Multicore HPC Applications. In: SIMPÓSIO EM SISTEMAS COMPUTACIONAIS DE ALTO DESEMPENHO (SSCAD), 27. , 2026, Natal/RN.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2026
.
p. 192-205.
DOI: https://doi.org/10.5753/sscad.2026.34906.
