Vol. 30, issue 11, article # 3

Abstract:

Issues about improving the performance of Monte Carlo numerical simulation of light transport in the Earth’s atmosphere by moving from consecutive calculations to parallel ones are discussed. A new parallel algorithm oriented to a computing system with a graphics processor that supports the NVIDIA CUDA technology is suggested. The efficiency of parallelization is analyzed on the basis of calculating the fluxes of downward and upward solar radiation in both vertically homogeneous and heterogeneous models of the atmosphere. The results of approbation of the new code under various atmospheric conditions including continuous single-layered and multilayered clouds and selective molecular absorption are presented. The results of testing the code using video cards with different compute capability are analyzed. It is shown that the changeover of computing from conventional PCs to the architecture of graphics processors gives more than a hundredfold gain in performance and fully reveals the capabilities of the technology used.

Keywords:

Monte Carlo method, solar radiation fluxes, parallel computing, GPU, CUDA technology, computation speedup

References:

1. Dubovik O., Lapyonok T., Litvinov P., Herman M., Fuertes D., Ducos F., Lopatin A., Chaikovsky A., Torres B., Derimian Y., Huang X., Aspetsberger M., Federspiel C. GRASP: A versatile algorithm for characterizing the atmosphere // SPIE: Newsroom. 2014. DOI: 10.1117/2.1201408.005558.
2. Lenobl' Zh. Perenos radiacii v rasseivajushhih i pogloshhajushhih atmosferah. L.: Gidrometeoizdat, 1990. 264 p.
3. Marchuk G.I., Mihajlov G.A., Nazaraliev M.A., Darbinjan R.A., Kargin B.A., Elepov B.S. Metod Monte-Karlo v atmosfernoj optike. Novosibirsk: Nauka, 1976. 280 p.
4. Marshak A., Davis A.B. 3D radiative transfer in cloudy atmospheres. Berlin: Springer, 2005. 686 p.
5. Moore G. Litography and the future of Moore’s law // Proc. SPIE. 1995. V. 2437. P. 2–17.
6. Kirkby D.R., Delpy D.T. Parallel operation of Monte Carlo simulations on a diverse network of computers // Phys. Med. Biol. 1997. V. 42, N 6. P. 1203–1208.
7. Colasanti A., Guida G., Kisslinger A., Liuzzi R., Quarto M., Riccio P., Roberti G., Villani F. Multiple processor version of a Monte Carlo code for photon transport in turbid media // Comput. Phys. Commun. 2000. V. 132, N 1–2. P. 84–93.
8. Kozhevnikova A.V., Tarasenkov M.V., Belov V.V. Parallel'nye vychislenija pri reshenii zadach vosstanovlenija kojefficienta otrazhenija zemnoj poverhnosti po sputnikovym dannym // Optika atmosf. i okeana. 2013. V. 26, N 2. P. 172–174; Kоzhevnikovа А.V., Таrаsеnkov М.V., Bеlоv V.V. Parallel computations for solving problem of the reconstruction of the reflection coefficient of the Earth’s surface by satellite data // Atmos. Ocean. Opt. 2013. V. 26, N 4. P. 326–328.
9. Glinskij B.M., Rodionov A.S., Marchenko M.A., Podkorytov D.I., Vins D.V. Agentno-orientirovannyj podhod k imitacionnomu modelirovaniju super JeVM jekzaflopsnoj proizvoditel'nosti v prilozhenii k raspredelennomu statisticheskomu modelirovaniju // Vestn. JuUrGU. 2012. N 18(277), iss. 12. P. 94–99.
10. Volkov K.N., Derjugin Ju.N., Emel'janov V.N., Karpenko A.G., Kozelkov A.S., Teterina I.V. Metody uskorenija gazodinamicheskih raschetov na nestrukturirovannyh setkah. M.: FIZMATLIT, 2014. 536 p.
11. Boreskov A.V., Harlamov A.A., Markovskij N.D., Mikushin D.N., Mortikov E.V., Myl'cev A.A., Saharnyh N.A., Frolov V.A. Parallel'nye vychislenija na GPU: arhitektura i programmnaja model' CUDA. M.: Izd-vo Mosk. un-ta, 2012. 336 p.
12. Zhu C., Liu Q. Review of Monte Carlo modeling of light transport in tissues // J. Biomed. Opt. 2013. V. 18, N 5. P. 050902-1–050902-12.
13. Fiks I.I. Ispol'zovanie graficheskih processorov dlja reshenija zadachi rasprostranenija sveta v diffuzionnoj fluorescentnoj tomografii metodom Monte-Karlo // Vestn. Nizhegorodsk. un-ta im. N.I. Lobachevskogo. 2011. N 4(1). P. 190–195.
14. Kirillin M.Ju., Fiks I.I., Katichev A.R., Gorshkov A.V., Gergel' V.P. Vysokoproizvoditel'nye vychislenija dlja zadach opticheskoj biomedicinskoj diagnostiki // Superkomp'juternye tehnologii v nauke, obrazovanii i promyshlennosti / pod red. V.A. Sadovnichego, G.I. Savina, Vl.V. Voevodina. Iss. 3. M.: Izd-vo Mosk. un-ta, 2012. 232 p.
15. Petrov D.A. Modelirovanie opticheskih metodov biomedicinskoj diagnostiki // Koncept. 2016. V. 11. P. 2851–2855.
16. Alerstam E., Lo W.C.Y., Han T.D., Rose J., Anderson-Engels S., Lilge L. Next-generation acceleration and code optimization for light transport in turbid media using GPUs // Biomed. Opt. Express. 2010. V. 1, N 2. P. 658–675.
17. Alerstam E., Svensson T., Anderson-Engels S. Parallel computing with graphics processing units for high-speed Monte Carlo simulation of photon migration // J. Biomed. Opt. 2008. V. 13, N 6. P. 060504-1–060504-3.
18. Cai F., He S. Using graphics processing units to accelerate perturbation Monte Carlo simulation in a turbid medium // J. Biomed. Opt. 2012. V. 17, N 4. P. 040502-1–040502-3.
19. Martinsen P., Blaschke J., Künnenmeyer R., Jordan R. Accelerating Monte Carlo simulations with an NVIDIA graphics processor // Comput. Phys. Commun. 2009. V. 180, N 10. P. 1983–1989.
20. Fang Q., Boas D.A. Monte Carlo simulation of photon migration in 3D turbid media accelerated by graphics processing units // Opt. Express. 2009. V. 17, N 22. P. 20178–20190.
21. Ren N., Liang J., Qu X., Li J., Lu B., Tian J. GPU-based Monte Carlo simulation for light propagation in complex heterogeneous tissues // Opt. Express. 2010. V. 18, N 7. P. 6811–6823.
22. Efremenko D.S., Loyola D.G., Doicu A., Spurr R.J.D. Multi-core-CPU and GPU-accelerated radiative transfer models based on the discrete ordinate method // Comput. Phys. Commun. 2014. V. 185, N 12. P. 3079–3089.
23. Ramon D., Steinmetz F., Compiegne M., Jolivet D. Massively parallel Monte-Carlo radiative transfer code on a desktop PC. URL: http://www-loa.univ-lille1.fr/workshops/Trattoria-2015/documents/posters/Didier_Ramon.pdf (last access: 1.10.2017).
24. Clough S.A., Iacono M.J., Moncet J.L. Line-by-line calculations of atmospheric fluxes and cooling rates: Application to water vapor // J. Geophys. Res. D. 1992. V. 97. P. 16519–16535.
25. NVIDIA official page. URL: www.nvidia.ru (lass access: 01.10.2017).
26. Russkova T.V., Zhuravleva T.B. Optimizacija posledovatel'nogo programmnogo koda dlja modelirovanija perenosa solnechnogo izluchenija v vertikal'no-neodnorodnoj srede // Optika atmosf. i okeana. 2016. V. 29, N 10. P. 836–842; Russkоvа Т.V., Zhurаvlevа Т.B. Optimization of sequential code for simulation of solar radiation transfer in a vertically heterogeneous environment // Atmos. Ocean. Opt. 2017. V. 30, N 2. P. 169–175.
27. Frank-Kameneckij A.D. Modelirovanie traektorij nejtronov pri raschete reaktorov metodom Monte-Karlo. M.: Atomizdat, 1978. 93 p.
28. Sanders Dzh., Kjendrot Je. Tehnologija CUDA v primerah: vvedenie v programmirovanie graficheskih processorov. M.: DMK Press, 2015. 232 p.
29. Zhukovskij M.E., Uskov R.V. Matematicheskoe modelirovanie radiacionnoj jemissii jelektronov na gibridnyh superkomp'juterah // Vychislitel'nye metody i programmirovanie. 2012. V. 13. P. 271–279.
30. Press W.H., Teukolsky S.A., Vetterling W.T., Flannery B.P. Numerical Recipes in Fortran 77: The Art of Scientific Computing. Volume 1 of Fortran Numerical Recipes. Cambridge University Press, 1986. 1003 pp.
31. Marsaglia G. Random Number Generators // J. Mod. Appl. Stat. Methods. 2003. V. 2, N 1. P. 2–13.
32. Lee A., Yau C., Giles M.B., Doucet A., Holmes C.C. On the utility of graphic cards to performe massively parallel simulation of advanced Monte Carlo methods // J. Comput. Graph. Stat. 2010. V. 19, N 4. P. 769–789.
33. Marsaglia G. Diehard battery of tests of randomness. The Marsaglia random number CD ROM. Department of Statistics, Florida State University, 1995.
34. Miller G.L. Riemann’s hypothesis and tests for primarily // J. Comput. Syst. Sci. 1976. V. 13. P. 300–317.
35. Rabin M.O. Probabilistic algorithm for testing primarily // J. Number Theory. 1980. V. 20, N 1. P. 128–138.
36. Matsumoto M., Nishimura T. Mersenne twister: A 623-dimensionally equidistributed uniform pseudo-random number generator // ACM Trans. Model. Comput. Sim. 1998. V. 8, N 1. P. 3–30.
37. Matsumoto M., Nishimura T. Dynamic creation of pseudorandom number generators // Monte Carlo and Quasi-Monte Carlo Methods. Springer, 2000. P. 56–69.
38. Hess M., Koepke P., Schult I. Optical properties of aerosols and clouds: The software package OPAC // Bull. Am. Meteorol. Soc. 1998. V. 79, N 5. P. 831–844.
39. Komarov V.S., Lomakina N.Ja. Statisticheskie modeli pogranichnogo sloja atmosfery Zapadnoj Sibiri. Tomsk: Izd-vo IOA SO RAN, 2008. 222 p.
40. Mazin I.P., Hrgian A.H., Imjaninov I.M. Oblaka i oblachnaja atmosfera. L.: Gidrometizdat, 1989. 647 p.
41. Anderson G.P., Clough S.A., Kneizys F.X., Chetwynd J.H., Shettle E.P. AFGL Atmospheric Constituent Profiles (0–120 km). Air Force Geophysics Laboratory, 1986. 46 p.
42. Firsov K.M., Smirnov A.B. Predstavlenie funkcij propuskanija rjadom jeksponent // Optika atmosf. i okeana. 1995. V. 8, N 8. P. 1248–1252.
43. Tvorogov S.D. Nekotorye aspekty zadachi o predstavlenii funkcii pogloshhenija rjadom jeksponent // Optika atmosf. i okeana. 1994. V. 7, N 3. P. 315–326.
44. Fischer J., Grassl H. Detection of cloud top height from backscattered radiances within the oxygen A band. Part 1. Theoretical study // J. Appl. Meteorol. 1991. N 30. Р. 1245–1259.
45. Koelemeijer R.B.A., Stammes P., Hovenier J.W., de Haan J.F. A fast method for retrieval of cloud parameters using oxygen A band measurements from the Global Ozone Monitoring Experiment // J. Geophys. Res. 2001. V. 106. Р. 3475–3490.
46. Badaev V.V., Malkevich M.S., Nizik B., Cimmerman G. Opredelenie opticheskih parametrov zemnoj poverhnosti, okeana i atmosfery so sputnikov «Interkosmos 20 i 21» // Issled. Zemli iz kosmosa. 1985. N 5. P. 18–29.
47. Timofeyev Yu.M., Vasilyev A.V., Rozanov V.V. Information content of the spectral measurement of the 0.76 mm O₂ outgoing radiation with respect to the vertical aerosol optical properties // Adv. Space Res. 1995. V. 16, N 10. Р. 91–94.