Journal of Computational Neuroscience

, Volume 5, Issue 4, pp 443–459 | Cite as

Large Neural Network Simulations on Multiple Hardware Platforms

  • Per Hammarlund
  • Örjan Ekeberg


To efficiently simulate very large networks of interconnected neurons, particular consideration has to be given to the computer architecture being used. This article presents techniques for implementing simulators for large neural networks on a number of different computer architectures. The neuronal simulation task and the computer architectures of interest are first characterized, and the potential bottlenecks are highlighted. Then we describe the experience gained from adapting an existing simulator, SWIM, to two very different architectures–vector computers and multiprocessor workstations. This work lead to the implementation of a new simulation library, SPLIT, designed to allow efficient simulation of large networks on several architectures. Different computer architectures put different demands on the organization of both data structures and computations. Strict separation of such architecture considerations from the neuronal models and other simulation aspects makes it possible to construct both portable and extendible code.

computational neuroscience neural networks neural simulation large-scale simulation portable software 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. Agerwala T, Martin JL, Mirza JH, Sadler DC, Dias DM, and Snir M (1995) SP2 System Architecture. IBM Systems Journal 34: 152–184.Google Scholar
  2. Alverson R, Callahan D, Cummings D, Koblenz B, Porterfield A, and Smith B (1990) The Tera computer system. In: Proceedings of the 1990 ACM International Conference on Supercomputing. pp. 1–6.Google Scholar
  3. Amdahl GM (1967) The validity of the single processor approach to achieving large scale computing capabilities. In: AFIPS Conf. Proc. Spring Joint Comput. Conf 31, pp. 483–485.Google Scholar
  4. Andersson TE, Culler DE, Patterson DA, and the NOW team (1995) A case for NOW (Networks of Workstations). IEEE Micro 15: 54–64.Google Scholar
  5. Bailey DH (1993) RISC microprocessors and scientific computing. In: Proceedings, Supercomputing' 93: Portland, Oregon, November 15–19, 1993. ACM Press, New York, pp. 645–655.Google Scholar
  6. Bower JM and Beeman D (1995) The book of Genesis. TELOS. Springer Verlag, New York.Google Scholar
  7. De Schutter E (1992) A consumer guide to neuronal modeling software. Trends Neurosci. 15: 462–464.Google Scholar
  8. Destexhe A, Mainen ZE, and Sejnowski TJ (1994) An efficient method for computing synaptic conductances based on a kinetic model of receptor binding. Neural Comp. 6: 14–18.Google Scholar
  9. Ekeberg Ö, Hammarlund P, Levin B, and Lansner A. (1994) SWIM: A simulation environment for realistic neural network modeling. In: Skrzypek, J, ed. Neural Network Simulation Environments, Kluwer Academic Publishers, Boston, MA pp. 47–71.Google Scholar
  10. Foster I (1995) Designing and Building Parallel Programs. Addison-Wesley Publishing Company.Google Scholar
  11. Fransén E and Lansner A (1990) Modelling Hebbian cell assemblies comprised of cortical neurons. In: Proc. Open Network Conference on Neural Mechanisms of Learning and Memory. London. pp. 4:9.Google Scholar
  12. Fujimoto R. M (1990) Parallel discrete event simulation. Communications of the ACM 33: 30–53.Google Scholar
  13. Golub G. H and van Loan C. F (1989) Matrix Computations (2nd edition). Johns Hopkins University Press, 2nd edition.Google Scholar
  14. Gropp W, Lusk E, and Skjellum A (1994) Using MPI Portable Parallel Programming with the Message-Passing Interface. MIT Press, Cambridge, Massachusetts.Google Scholar
  15. Hammarlund P (1996) Communication algorithms for networks with connection delays. In: Hammarlund P, ed. Techniques for Efficient Parallel Scientific Computing. Royal Institute of Technology, Stockholm. pp. 73–104. PhD thesis, TRITA-NA-P9611.Google Scholar
  16. Hammarlund P, Ekeberg Ö, Wilhelmsson T, and Lansner A. (1996) Large neural network simulations on multiple hardware platforms. In: Bower JM, eds. The Neurobiology of Computation: Proceedings of the Fifth Annual Computation and Neural Systems Conference. Plenum Press, Boston, MA.Google Scholar
  17. Hellgren J, Grillner S, and Lansner A. (1992) Computer simulation of the segmental neural network generating locomotion in lamprey by using populations of network interneurons. Biol. Cybern. 68: 1–13.Google Scholar
  18. Hennessy JL and Patterson DA. (1995) Computer Architecture: A Quantitative Approach. Morgan Kaufmann Publishers, Inc., 2nd edition.Google Scholar
  19. Hines M. (1984) Efficient computation of branched nerve equations. Int. J. Bio-Medical Computing 15: 69–76.Google Scholar
  20. Hines M (1993a) NEURON-a program for simulation of nerve equations. In: Eeckman F, ed. Neural Systems: Analysis and Modeling. Kluwer Academic Publishers, Boston, MA. pp. 127–136.Google Scholar
  21. Hines M (1993b) The NEURON simulation program. In Skrzypek, ed. Neural Network Simulation Environments. Kluwer Academic Publishers, Boston, MA.Google Scholar
  22. Hodgkin AL and Huxley AF (1952) A quantitative description of membrane current and its application to conduction and excitation in nerve. J. Physiol. 117: 500–544.Google Scholar
  23. Koch C and Segev I (1989) Methods in neuronal modeling: from synapses to networks. In: Computational Neuroscience. MIT Press, Cambridge, MA.Google Scholar
  24. Kohn MC, Hines M, Kootsey JM, and Feezor MD (1989) A block organized model builder. In: Proceedings of the Symposium on Physilogical Modeling at the 7th ICCM. Chicago.Google Scholar
  25. Lansner A and Fransén E (1995a) Cortical columns in attractor models makes connectivity more biologically realistic. In: Abeles M and Sompolinsky H, eds. Cortical Dynamics in Jerusalem: Proceedings of the Symposium on Experimental and Theoretical Issues in the Dynamics and Function of the Neocortex. The Hebrew University of Jerusalem, Jerusalem, Israel pp. 89.Google Scholar
  26. Lansner A and Fransén E (1995b) Improving the realism of attractor models by using cortical columns as functional units. In: Bower JM, ed. The Neurobiology of Computation: Proceedings of the Third Annual Computation and Neural Systems Conference. Kluwer Academic Publishers, Boston, MA pp. 251–256.Google Scholar
  27. Mascagni MV (1989) Numerical methods for neuronal modeling. In: Methods in Neuronal Modeling: From Synapses to Networks. (Koch and Segev, 1989), pp. 439–484.Google Scholar
  28. Nelson ME, Furmanski W, and Bower JM (1989) Simulating neurons and networks on parallel computers. In: Methods in Neuronal Modeling: From Synapses to Networks. (Koch and Segev, 1989) pp. 397–437.Google Scholar
  29. Rall W (1977) Core conductor theory and cable properties of neurons. In: Brookhart M, Mountcastle V, and Kandel E, eds. Handbook of Physiology (Sect. 1). The Nervous System. I. Cellular Biology of Neurons. American Physiological Society, Bethesda, MD pp. 39–97.Google Scholar
  30. Righter R and Walrand JC (1989) Distributed simulation of discrete event systems. Proc. of the IEEE 77: 99–113.Google Scholar
  31. Sunderam VS, Geist GA, Dongarra J, and Manchek R (1994) The PVM concurrent computing system: Evolution, experiences, and trends. Parallel Computing 20: 531–546.Google Scholar
  32. Tråvén H, Brodin L, Lansner A, Ekeberg Ö, Wallén P, and Grillner S (1993) Computer simulations of NMDA and non-NMDA receptor-mediated synaptic drive—sensory and supraspinal modulation of neurons and small networks. J. Neurophysiol. 70: 695–709.Google Scholar
  33. Wadden T, Hellgren-Kotaleski J, Lansner A, and Grillner S (1995) Simulations of intersegmental coordination using a continuous network model. In: Bower JM, ed. The Neurobiology of Computation: Proceedings of the Third Annual Computation and Neural Systems Conference. Kluwer Academic Publishers, Boston, MA.Google Scholar
  34. Wallén P, Ekeberg Ö, Lansner A, Brodin L, Tråvén H, and Grillner S (1992) A computer-based model for realistic simulations of neural networks. II: The segmental network generating locomotor rhythmicity in the lamprey. J. Neurophysiol. 68: 1939–1950.Google Scholar
  35. Wilson MA and Bower JM (1989) The simulation of large-scale neural networks. In: Methods in Neuronal Modeling: From Synapses to Networks. (Koch and Segev, 1989) pp. 291–333.Google Scholar

Copyright information

© Kluwer Academic Publishers 1998

Authors and Affiliations

  • Per Hammarlund
    • 1
  • Örjan Ekeberg
    • 1
  1. 1.Studies of Artificial Neural Systems (SANS), Department of Numerical Analysis and Computing ScienceRoyal Institute of TechnologyStockholmSweden

Personalised recommendations