2014-№1(42) Article 17

V.V. Lukashenko, V.A. Romanchuk
Mathematical modeling of virtual clusters of grid-enabled system programs on the basis of neuroprocessors. P. 175-182.

UDC 004.75, 004.272

The paper provides a survey of mathematical modeling of virtual clusters reconfigured for some specific tasks. Production nodes of such clusters function on the basis of neuroprocessors. The paper also provides detailed models of executive programs and control modules for certain nodes.

Distributed computing, neuroprocesses, clusterization.

References:

1. Foster, I. The anatomy of the GRID [Text] // I. Foster, C. Kesselman, St. Tuecke // International Journal of High Performance Computing Applications. – 2001. – No 3/15. – P. 200–222.

2. Lublin, U. The workload on parallel supercomputers: modeling the characteristics of rigid job [Text] / U. Lublin, G. Feitelson // Journal of Parallel and Distributed Computing archive. – 2003. – Vol. 63. – Iss. 11. – P. 542–546.

3. Grigorenko, D. Improving recoverability cluster neuroprocessor data processing systems [Text] / D. Grigorenko, V. Ruchkin // Digital signal processing. – 2012. – No 3. – P. 67–72.

4. Polezhaev, P. Experimental study of scheduling algorithms for computational grid problems-system [Text] // Control Systems and Information Technology. – 2011. – No3.2/45. – P. 266–270.