LABORATORY OF PARALLEL AND DISTRIBUTED SUPERCOMPUTING TECHNOLOGIES

Head of the Laboratory -
Dr.Sc. (Phys.-Math.), Professor E.A. Nurminsky,
Honored Scientist of the Russian Federation.

Staff

Total-20,
Researchers - 8,
Dr. Sc. - 1,
PhD - 5.

http://www.iacp.dvo.ru/lab_11/


Main research directions

  • Elaboration of the computational methods for solving complex decision-making problems, mathematical modeling using massive parallel large-powered computational complexes.
  • Development of technology of parallel programming based on the Petri networks formalism.
  • Development of network infrastructure of FEB RAS and network information services.
  • Development of the Center of Multiple Access "Far Eastern Computing Resources".

Principal results

  • The methods of direct-dual decomposition, parallel projection methods, algorithms of uneven optimization with memory restrictions, methods for solving variant inequalities have been elaborated.
  • Software to specify and analyse parallel programs in the terms of Petri network has been developed.
  • The models and methods for solving the problems of mathematical modeling of supersonic gas flows with energy-release have been developed.

Representative publications

  1. Golenkov E.A., Sokolov A.S. Method of automatic model building of parallel program in the terms of Petri network. Computational Methods and Programming. V. 6, № 2, Issue of Moscow Univ., 2005. pp. 220-229.
  2. Levin V.A., Annenkov V.A., Trifonov E.V. Destruction of incident shock wave by energy-release source. Applied Mechanics and Technical Physics., Issue 2, 2006.
  3. Velichko A.S., Nurminsky E.A. Direct-dual decomposition of the problem about replication of market assets portfolio. Autovation and Telemechanics,2, 2004. pp. 170-178.
  4. Nurminsky E.A. Convergence of suitable affine spaces method for solving the problem about the smallest distance up to simplex. Journal of Computational Mathematics and Mathematical Physics. V. 45, Issue 11, 2005. pp. 1996-2004.
  5. Nurminsky E.A. Method of separation planes with restricted memory for solving the problems of convex uneven optimization. Computational Methods and Programming. V. 7, № 1, Issue of Moscow Univ., 2006. pp. 133-137.