Cache Coherence Protocol Overhead in NUMA-Aware Distributed Shared Memory Systems: An Empirical Analysis of MOESI State Transitions Under Heterogeneous Workload Contention

Authors

  • Riley Wright Professor
  • Chris Gonzalez Associate Professor
  • Drew Anderson PhD
  • Jamie Robinson D.Sc

Keywords:

NUMA architecture, MOESI cache coherence protocol, distributed shared memory, false sharing, directory-based coherence, hardware performance counters, memory hierarchy optimization, inter-node invalidation latency, heterogeneous workload contention

Abstract

Cache coherence maintenance in Non-Uniform Memory Access (NUMA) architectures imposes measurable latency penalties during high-frequency MOESI state transitions, particularly under asymmetric workload distributions across heterogeneous compute nodes. This study presents a rigorous empirical analysis of protocol-level overhead in distributed shared memory (DSM) systems comprising 32-node clusters with mixed CPU-GPU memory hierarchies. Using hardware performance counters, cycle-accurate simulation via gem5, and custom microbenchmark suites, we quantify inter-node invalidation traffic, false-sharing penalties, and directory-based coherence arbitration costs across six distinct contention scenarios. Results demonstrate that MOESI transition storms account for up to 34.7% of total execution stall cycles in write-intensive workloads, with cross-socket invalidation latencies reaching 412 ns under peak contention. Proposed adaptive ownership migration heuristics reduce coherence overhead by 21.3% without sacrificing memory consistency guarantees.

Author Biographies

Riley Wright, Professor

Professor
Technische Universität München
Arcisstraße 21, 80333 Munich, Bavaria, Germany

Chris Gonzalez, Associate Professor

Associate Professor
Korea Advanced Institute of Science and Technology (KAIST)
291 Daehak-ro, Yuseong-gu, Daejeon 34141, Republic of Korea

Drew Anderson, PhD

PhD
University of Toronto
27 King's College Circle, Toronto, Ontario M5S 1A1, Canada

Jamie Robinson, D.Sc

D.Sc
Delft University of Technology
Mekelweg 2, 2628 CD Delft, South Holland, Netherlands

References

Kumar, Nitin, and Vipin Kataria. "Enhanced Sentiment Classification using a Multi-layered Stacked Ensemble Architecture."

Kumar, N., & Kataria, V. (2023). Enhanced Sentiment Classification using a Multi-layered Stacked Ensemble Architecture. International Journal of Intelligent Systems and Applications in Engineering, 11(4s), 304–311.

D. Satyanarayana, Sathyashree. S “Mobility Management in Voronoi based Cell Structure for Wireless networks”, IEEE proceedings for the International Conferences on Currents trends in Information Technology (CTIT), Pages: 204-208, Dubai, December-2013. DOI: 10.1109/CTIT.2013.6749504; INSPEC Accession Number: 14131032

D. Satyanarayana and J. M. H. Elmirghani, "An Energy Efficient Network Architecture for Infrastructured Wireless Networks," 2010 IEEE Global Telecommunications Conference GLOBECOM 2010, Miami, FL, USA, 2010, pp. 1-6, doi: 10.1109/GLOCOM.2010.5683234.

D. Satyanarayana and J. M. H. Elmirghani, "A Voronoi Based Energy Efficient Architecture for Wireless Networks," 200S Third International Conference on Next Generation Mobile Applications, Services and Technologies, Cardiff, UK, 2009, pp. 377-382, doi: 10.1109/NGMAST.2009.79.

D. Satyanarayana and S. V. Rao, "Local Delaunay Triangulation for Mobile Nodes," 2008 First International Conference on Emerging Trends in Engineering and Technology, Nagpur, India, 2008, pp. 282-287, doi: 10.1109/ICETET.2008.253.

D. Satyanarayana, S. Chattopadhyay and J. Sasidhar, "Low power combinational circuit synthesis targeting multiplexer-based FPGAs," 17th International Conference on VLSI Design. Proceedings., Mumbai, India, 2004, pp. 79-84, doi: 10.1109/ICVD.2004.1260906.

Alkalbani, A.S., Mantoro, T., Degala, S. (2016). Energy-Distance Aware Clustering Scheme (E-DACS) for Wireless Sensor Networks. In: Meesad, P., Boonkrong, S., Unger, H. (eds) Recent Advances in Information and Communication Technology 2016. Advances in Intelligent Systems and Computing, vol 463. Springer, Cham. https://doi.org/10.1007/978-3-319-40415-8_27

Satyanarayana, D., Sathyashree, A. A. K., & Alkalbani, A. (2015, November). Voronoi LEACH for Energy Efficient Communication in Wireless Sensor Networks. In Proceedings for the International Conference on Electronics Computer networks and Information Technology, Dubai.

Гашимов, А. М., Гулиев, Г. Б., & Рахманов, Н. Р. (2014). Улучшенный алгоритм нечеткой логики для управления реактивной мощностью и напряжением в распределительных сетях. Энергетика. Известия высших учебных заведений и энергетических объединений СНГ, (2), 29-39.

Yusifbayli, N. A., & Guliyev, H. B. (2011). Intelligent control system of voltage regimes in the electrical networks. IEEE PES innovative smart grid Europe, 5-7.

Rahmanov, N., Guliyev, H., İbtahimov, F., & Mammadov, Z. (2024). Determination of optimal dimensions of hybrid AC/DC distributed generation system with renewable sources for autonomous power supply of remote locations. In E3S Web of Conferences (Vol. 584, p. 01020). EDP Sciences.

Guliyev, H. B. (2024). Management modes of reactive power compensation facilities in networks with renewable energy sources with distorting loads. IJTPE Journal International Journal on Technical and physical problems of engineering (IJTPE), 16(1), 14-20.

Hashimov, A. M., Babayeva, A. R., & Guliyev, H. B. (2019). Shunt reactors control algorithm using fuzzy sets theory. International Journal on Technical and Physical Problems of Engineering, 11(1), 10-15.

Published

2024-10-24

Issue

Section

Articles