Computer simulation is an important tool for studying epidemic dynamics. Owing to the scales and runtime speed requirements, the simulations of large-scale pandemic diseases such as severe acute respiratory syndrome (SARS) and H1N1 influenza usually require high-performance parallel simulation or computation. Previous works on the parallel simulations of large-scale epidemic were implemented on traditional general purpose CPU-based platforms or clusters. As more and more high-performance computation clusters are being built with so-called general-purpose graphics cards, this paper presents the implementation and optimization techniques for social contact network-based large-scale epidemic simulation on GPU clusters. Compared with previous works, this paper focuses on (1) how to efficiently implement the contact network-based parallel epidemic simulation on GPUs, and (2) how to hide communication latencies between processing nodes to improve scalability. Our proposed techniques are implemented and tested on a commodity cluster whose processing nodes are equipped with GPGPUs. The experimental results show that, for the simulation of 20 million individuals and 1.2 billion host contacts on 80 nodes, the execution on GPUs can achieve 7.4×−11.7× speedup over the execution on CPUs.
BarrettCLBissetKREubankSGFengXMaratheMV. EpiSimdemics: an efficient algorithm for simulating the spread of infectious disease over large realistic social networks. In Proceedings of the 2008 ACM/IEEE International Conference for High Performance Computing, Networking, Storage and Analysis, 2008, paper No. 37, pp. 1–12.
2.
EubankS. Scalable, efficient epidemiological simulation. In Proceedings of the 2002 ACM Symposium on Applied Computing, 2002, pp. 139–145.
3.
EubankSGucluHMaratheMV. Modelling disease outbreaks in realistic urban social networks. Nature2004; 429: 180–184.
4.
HalloranMEFergusonNMEubankS. Modeling targeted layered containment of an influenza pandemic in the United States. Proc Natl Acad Sci USA2008; 105: 4639–4644.
5.
CarleyKMFridsmaDBCasmanE. Biowar: Scalable agent-based model of bioattacks. IEEE Trans Syst Man Cybernet A2006; 36: 252–265.
6.
BarrettCLEubankSGSmithJP. If smallpox strikes Portland. Scientific American2003; 292(3): 42–49.
7.
HsiehJLHuangCYSunCTChenYMA. Using the CAMIM small-world epidemic model to analyze public health policies. In Proceedings of Western Multiconference on Health Sciences Simulation, 2005, pp. 63–69.
8.
HuangCYSunCTHsiehJLLinH. Simulating SARS: small-world epidemiological modeling and public health policy assessments. J Artif Soc Social Sim2004; 7(4). http://jasss.soc.surrey.ac.uk/7/4/2.html.
9.
HuangCYSunCTHsiehJLChenYMALinH. A novel small-world model: using social mirror identities for epidemic simulations. Simulation2005; 81: 671–699.
10.
BarthelemyMBarratAPastor-SatorrasRVespignaniA. Velocity and hierarchical spread of epidemic outbreaks in scale-free networks. Phys Rev Lett2004; 92: 178701-178704.
WattsDJStrogatzSH. Collective dynamics of small-world networks. Nature1998; 393: 440–442.
14.
BissetKRChenJFengXKumarVSAMaratheMV. EpiFast: a fast algorithm for large scale realistic epidemic simulations on distributed memory systems. In Proceedings of the 2009 ACM/IEEE International Conference for High Performance Computing, Networking, Storage and Analysis, 2009, pp. 430–439.
15.
PerumallaKSSealSK. Reversible parallel discrete-event execution of large-scale epidemic outbreak models. In Proceedings of the 2010 IEEE Workshop on Principles of Advanced and Distributed Simulation, 2010, pp. 1–8.
16.
HuangCYTsaiYSWenTH. A network-based simulation architecture for studying epidemic dynamics. Simulation2010; 86: 351–368.
17.
MeyersLA. Contact network epidemiology: bond percolation applied to infectious disease prediction and control. Bull Amer Math Soc2007; 44: 63–86.
MeyersLAPourbohloulBNewmanMEJSkowronskicDMBrunhamRC. Network theory and SARS: predicting outbreak diversity. J Theoret Biol2005; 232: 71–81.
20.
AabyBGPerumallaKSSealSK. Efficient simulation of agent-based models on multi-GPU and multi-core clusters. In Proceedings of the 3rd International ICST Conference on Simulation Tools and Techniques, 2010, article 29.
21.
FengZHuangWCastillo-ChavezC. Global behavior of a multi-group SIS epidemic model with age structure. J Differential Eqs2005; 218: 292–324.
22.
InabaH. Age-structured homogeneous epidemic systems with application to the MSEIR epidemic model. J Math Biol2007; 54: 101–146.
23.
LanglaisMNaulinJM. An age structured SI epidemic problem in a heterogeneous environment. Evol Equat Appl Phys Ind Life Sci Econ2003; 55: 307–321.
24.
ShimEFengZMartchevaMCastillo-ChavezC. An age-structured epidemic model of rotavirus with vaccination. J Math Biol2006; 53: 719–746.
25.
SupriatnaAKSoewonoEVan GilsSA. A two-age-classes dengue transmission model. Math Biosci2008; 216: 114–121.
26.
WangWZhaoXQ. An age-structured epidemic model in a patchy environment. SIAM J Appl Maths2005; 65: 1597–1614.
27.
BaileyN. The Mathematical Theory of Infectious Diseases and Its Applications. New York: Hafner Press, 1975.
28.
KuznetsovYPiccardiC. Bifurcation analysis of periodic SEIR and SIR epidemic models. J Math Biol1994; 32: 109–121.
29.
MurrayJD. Mathematical Biology: I. An Introduction, 3rd edn.Berlin: Springer-Verlag, 2002.
30.
GermannTCKadauKLonginiIMMackenCA. Mitigation strategies for pandemic influenza in the United States. Proc Natl Acad Sci USA2006; 103: 5935–5940.
31.
ParkerJ. A flexible, large-scale, distributed agent based epidemic model. In Proceedings of the 39th Conference on Winter Simulation, 2007, pp. 1543–1547.
32.
LinebargerJMGoldsbyMEFelligDHawleyMFMoorePCSaTJ. Smallpox over San Diego: joint real-time federations of distributed simulations and simulation users under a common scenario. In Proceedings of the Workshop on Parallel and Distributed Simulation, 2007, pp. 7–14.
33.
PerumallaKSAabyBG. Data parallel execution challenges and runtime performance of agent simulations on GPUs. In Proceedings of the 2008 Spring Simulation Conference, 2008, pp. 116–123.
34.
D’SouzaRMLysenkoMRahmaniK. SugarScape on steroids: simulating over a million agents at interactive rates. In Proceedings of Agent 2007 Conference on Complex Interaction and Social Emergence, 2007.
