Empirical Analysis of Spiking Neural Networks Using CPU, GPU and GPU Clusters
Sreenivasa.N1, S. Balaji2
1Sreenivasa.N, Research Scholar-Jain University, Dept. of Computer Science & Engineering Nitte Meenakshi Institute of Technology, Yelahanka Bengaluru-560064, India.
2S. Balaji, Centre for Incubation, Innovation, Research and Consultancy Jyothy Institute of Technology, Tataguni, Bengaluru-560082, India.
Manuscript received on 18 June 2019 | Revised Manuscript received on 25 June 2019 | Manuscript published on 30 June 2019 | PP: 1942-1950 | Volume-8 Issue-5, June 2019 | Retrieval Number: E7886068519/19©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Many attempts have been made to study the neural networks and to model them. These attempts have led to the development of neural network simulation software packages such as GENESIS and NEURON which have been the de-facto simulators for some time now. However, further studies have found that one of the major hindrances in using the aforementioned simulators is speed. These simulators use time driven technique which isolates the mimicked time to brief time periods and in every progression the factors of neural states are estimated and reiterated through a numerical examination strategy. This method includes complex calculations which do not foster the development of scalable neural systems; hence, there is a demand for quick re-enactment of neural systems which led to an alternative strategy known as event driven simulation. This technique processes and appraises the neural state factors when another event alters the typical advancement of the neuron, that is, the point at which information is created. In the meantime, it is realized that the data communication in neural networks is done by the purported spikes. Less than 1% of the neurons are at the same time dynamic which catalysis the efficiency of eventdriven Spiking Neural Networks (SNN) simulation. Brain is one of the most complex human organs. For long this has intrigued several researchers from various disciplines in the world. A lot of research has been reported on brain since early days especially from the physiological and psychological angles. However, the advances in the computing technologies especially the High Performance Computing (HPC) platform has opened several new avenues for researchers to carry out their research. In the insilico approach of modelling the brain, the simulation is performed by constructing the networks which draw inspiration from the simple neuron model. We present an empirical study on the hybrid CPU-GPU model-based simulation of SNN which are based on our works .
Keywords: SNN, CPU-GPU Co-Processing, High Performance Computing, Neural Networks, Numerical Analysis.
Scope of the Article: High Performance Computing