Founder: Northern (Arctic) Federal University named after M.V. Lomonosov

Editorial office address: Russian Federation, 163002, Arkhangelsk, Naberezhnaya Severnoy Dviny 17, office 1410a

Phone: (818-2) 21-61-00(15-33)
e-mail: l.zhgileva@narfu.ru
http://aer.narfu.ru/en/

16+

ABOUT

Fine-Grained Local Parallel Algorithms. Pp. 110–115.

 Версия для печати

Section: Physics. Mathematics. Informatics

UDC

[004.421.2+004.9](045)

Authors

Vorobyev Vladimir Anatolyevich, Institute of Mathematics, Information and Space Technologies, Northern (Arctic) Federal University named after M.V. Lomonosov (Arkhangelsk, Russia) 

Yufryakova Olga Alekseevna, Institute of Mathematics, Information and Space Technologies, Northern (Arctic) Federal University named after M.V. Lomonosov (Arkhangelsk, Russia)

Abstract

The paper examines the concept of fine-grained local parallel algorithm and states its properties. Two examples of original MLP-algorithms on the ring and toroidal computing structures are given. Complexity and efficiency of parallelization were estimated.

Keywords

computational complexity, algorithm structure, parallel computing, massive parallelism, finegrained local parallel algorithms, locality, architecture of a parallel computing system, computational speed-up.

The full-text version of the article can be requested through the university’s library.

References

  1. Voevodin V.V. Vychislitel’naya matematika i struktura algoritmov [Computational Mathematics and Algorithm Structure]. Moscow, 2010. 168 p. 
  2. Vorob’ev V.A. Ob effektivnosti parallel’nykh vychisleniy [On the Efficiency of Parallel Computing]. Avtometriya, 2000, no 1, pp. 50–58. 
  3. Derbina Yu.V., Berezovskiy V.V. Podkhody k realizatsii melkozernistoy lokal’no-parallel’noy arkhitektury dlya modelirovaniya ekologicheskikh riskov [Approaches to Implementation of Fine-grained Local-parallel Architecture for Environmental Risks Modeling]. Materialy IX mezhdunarodnoy konferentsii-seminara «Vysokoproizvoditel’nye parallel’nye vychisleniya na klasternykh sistemakh» [Proc. 9th int. conf.-seminar “High-performance parallel computing on cluster systems”]. Vladimir, 2009, pp. 140–144. 
  4. Yufryakova O.A. Optimizatsiya kolichestva protsessorov dlya effektivnogo ispolneniya parallel’nykh algoritmov [Optimization of the Number of Processors for Efficient Execution of Parallel Algorithms]. Nauchnyy servis v seti Internet: poisk novykh resheniy: Trudy Mezhdunarodnoy superkomp’yuternoy konferentsii. [Scientific service in the Internet: exaflop future: Proc. int. supercomputing conf.]. Moscow, 2012. 
  5. Yufryakova O.A. Prepyatstviya k neogranichennomu masshtabirovaniyu parallel’nykh system [Obstacles to Unlimited Scalability of Parallel Systems]. Nauchnyy servis v seti Internet: ekzaflopsnoe budushchee: Trudy Mezhdunarodnoy superkomp’yuternoy konferentsii [Scientific service in the Internet: exaflop future: Proc. int. supercomputing conf.]. Moscow, 2011.