PROGRAM SYSTEMS: THEORY AND APPLICATIONS

12+

 

Online Scientific Journal published by the Ailamazyan Program Systems Institute of the Russian Academy of Sciences

2014 Issue 1
2014 Issue 2
2014 Issue 3
2014 Issue 4
2014 Issue 5

Papers are accepted in the form of a PDF file

To view the PDF files, you will need Adobe Acrobat Reader

    


• Содержание выпуска

Software and Hardware for Distributed Systems and Supercomputers

Responsible for the Section: Sergei Abramov, Dr. Phys.-Math.Sci., corresponding member of RAS

On the left: assigned number of the paper, submission date, the number of A5 pages contained in the paper,
and the reference to the full-text PDF .

 

Article # 1_2014

12 p.

PDF

submitted on 02th Dec 2013 displayed on website on 03th Feb 2014

Chudanov V.V., Goreinov S.A., Aksenova A.E., Pervichko V.A., Makarevich A.A.
New method for solving of CFD problems at clustered computers petascale performance

A new method for solving of CFD problems on the clustered petascale computers is described. The method is highly effective and can achieve significant performance under direct realization. A scalability this method close to ideal was demonstrated. (In Russian.)

Key words: CFD, clustered computers, petascale performance.

article citation

http://psta.psiras.ru/read/psta2014_1_3-14.pdf

Article # 2_2014

21 p.

PDF

submitted on 02th Dec 2013 displayed on website on 03th Feb 2014

Klimov A.V., Levchenko N.N., Okunev A.S., Stempkovsky A.L.
Supercomputers, memory hierarchy and dataflow computation model

Modern supercomputers are hierarchical, and the hierarchy depth tends to grow. Structure hierarchy (core – chip – node – card – cabinet – system) implies significant differences in communication time. Memory hierarchy also induces differences in access time: the larger is the size of the level, the slower is the access. Data movement overhead become the most significant factor of inefficiency, and thus the task of optimizing programs in this respect gets more and more difficult. We claim that these difficulties are largely a consequence of traditional programming paradigm that goes back to von Neumann. And although it has such a strong case as the acquirements and the legacy software, it is still useful at least in theory to understand the alternatives. We believe that the problem of the von Neumann programming model arises due to its exercising the so-called gather paradigm, as opposed to the scatter paradigm inherent to the proposed dataflow computation model, which provides more efficient and easier solution to the data movement optimization problem. (in Russian)

Key words: supercomputer, memory hierarchy, data preloading, dataflow computation model, gather paradigm, scatter paradigm, computation scheduling.

article citation

http://psta.psiras.ru/read/psta2014_1_15-36.pdf

Article # 3_2014

37 p.

PDF

submitted on 02th Dec 2013 displayed on website on 03th Feb 2014

Zatuliveter Yu.S., Fishchenko E.A., Artamonov S.E., Kozlov V.A.
Elements of the strategy of outstripping development and architectural prerequisites for the creation of single-chip accelerator for general-purpose massively-parallel computations based on the architecture of domestic multiprocessor computer PS-2000

The causes, manifestations and industrial problems of structural crisis of microprocessor architectures are analyzed. The state of the market of single-chip computer-accelerators for general-purpose massively parallel tasks is considered. The concept of further development of the domestic highly parallel computers PS-2000, which can be regarded as the first representatives of public multiprocessor systems of GP (General Purpose) class, is proposed. The various features of PS-2000M multiprocessor scalable architecture, as the successor of PS-2000, which possesses of great potential for outstripping development, are considered. A comparison of the PS-2000M architecture with modern single-chip architectures for massively parallel computers is given. (In Russian)

Key words: computer crisis, a massively parallel architecture, single-chip computer-accelerators, general purpose massively parallel computing, scalability, strategy of out-stripping development.

article citation

http://psta.psiras.ru/read/psta2014_1_37-74.pdf

Article # 4_2014

16 p.

PDF

submitted on 02th Dec 2013 displayed on website on 10th Feb 2014

Chudinov S.M., Malikov S.N.
Unified electronic modules — universal tool for high-level resources integration for the purposes of scientific and engineering missions

The article represents the structure, composition and sphere of application of the unified electronic modules (UEM products) to be used for the purposes of real time scale digital location with usage of the modified advanced algorithms of digital signal processing (DSP), as well as the design of scalable computer systems for digital data collecting and processing based on UEM for the purposes of digital location in various areas of science and technology. (in Russian.)

Key words: digital location, product UEM, FPGAs, MPT modules, submodules standard FMC, software, image processing in medicine, geophysics, and services of the Civil Defense Ministry.

article citation

http://psta.psiras.ru/read/psta2014_1_75-90.pdf

Article # 5_2014

20 p.

PDF

submitted on 02th Dec 2013 displayed on website on 17th Feb 2014

Besedin K.Y., Kostenetskiy P.S.
Simulating of query processing on multiprocessor database systems with modern coprocessors

This paper focuses on evaluation of database multiprocessor architectures with manycore coprocessors and GPUs. We implemented the emulator of parallel DBMS that uses computing cluster with NVIDIA GPUs or Intel Xeon Phi coprocessors for relational query processing. A number of experiments have been done using this emulator. (in Russian).

Key words: parallel DBMS, GPU, CUDA, Intel MIC, Intel Xeon Phi.

article citation

http://psta.psiras.ru/read/psta2014_1_91-110.pdf

Article # 6_2014

42 p.

PDF

submitted on 02th Dec 2013 displayed on website on 13th Feb 2014

Stegailov V.V., Norman G.E.
Challenges to the supercomputer development in Russia: a HPC user perspective

Over the past decade, active government support accelerated development of supercomputer industry in Russia. Today, there are several large supercomputers of large performance, which solve a growing number of problems. Supercomputer education in Russia is maturing. At the same time, one can observe an unbalanced development with respect to the supercomputer architectures, and a lack of massive parallelism in the solution of applied problems.
The article briefly describes the main trends of how the supercomputer architecture and interconnect evolved since the 1990s. The main trends highlighted, which is winning in a competitive environment of increasing demand for high performance computing using a growing number of processors (cores) for a given problem. We discuss what could be done to put Russia into the same main trend.
Critical attitude of this article seeks in no way to underestimate the progress in the development of Russian supercomputer industry. This is an attempt to focus the attention of the community on the challenges we feel today using HPC in research that could be competitive in the international context. (in Russian).

Key words: interconnect topology, way to exaflops era, scalability of parallel algorithms, perspective architectures.

article citation

http://psta.psiras.ru/read/psta2014_1_111-152.pdf

Article # 7_2014

21 p.

PDF

submitted on 02th Dec 2013 displayed on website on 17th Feb 2014

Zatuliveter Yu.S., Fishchenko E.A.
Principles of formation of universal seamlessly programmable and cybersecure algorithmic space

The problems and principles of formation in the resources of global networks seamlessly and universally programmable distributed computing envi-ronment are considered. The limitations of present computing environments, which heterogeneity is the cause of the combinatorial complexity of the problems of expansion distributed computing systems and providing of cybersecure are analyzed. For transition to universal, seamlessly programmable and cybersecure algorithmic space of distributed computing the mathematically closed form representation of data and programs is used. The requirements for the new element base (VLSI) — a network computer with nonmicroprocessor architecture with built — in hardware kernel functions of operating systems, providing cybersecure access to data and network routing are formulated. (in Russian).

Key words: global networks, distributed computing, mathematical homogeneous algorithmic space, seamless programming, universal network computer, nonmicroprocessor architecture, cybersecurity, smart memory.

article citation http://psta.psiras.ru/read/psta2014_1_153-173.pdf

Article # 8_2014

8 p.

PDF

submitted on 02th Dec 2013 displayed on website on 17th Feb 2014

Rumiantsev A.S.
Optimizing the Execution Time of a Desktop Grid Project

A model of computation time of a project in a Desktop Grid is viewed, under the following restrictions: the number of nodes is stationary, it takes each task the same time equal to deadline to complete, under the condition of deadline violation the task is to be calculated again. An inequality connecting probability of a single error in calculation of a task, number of replicas and number of tasks is concluded that shows when replication is necessary. An inequality connecting number of replicas, probability of a single error and penalty for error is concluded, which shows when replication is profitable. (in Russian).

Key words: desktop Grid, stochastic modeling, execution time.

article citation http://psta.psiras.ru/read/psta2014_1_175-182.pdf

Article # 9_2014

8 p.

PDF

submitted on 02th Dec 2013 displayed on website on 21th Feb 2014

Ivashko E.E.
Enterprise Desktop Grids

In this paper we consider Enterprise Desktop Grids, their capabilities, features, characteristics and applications. There are also examples of software aimed to construct Enterprise Desktop Grids. (in Russian).

Key words: Desktop Grid, Enterprise Desktop Grid, distributed computing.

article citation

http://psta.psiras.ru/read/psta2014_1_183-190.pdf

Article # 10_2014

54 p.

PDF

submitted on 02th Dec 2013 displayed on website on 28th Feb 2014

Kuksin A.Y., Lankin A.V., Morozov I.V., Norman G.E., Orekhov N.D., Pisarev V.V., Smirnov G.S., Starikov S.V., Stegailov V.V., Timofeev A.V.
Predictive modeling and simulation of properties and multi-scale processes in materials science. Tasks for Exaflops-era supercomputers

The approach is developed which allows to find out the problems which need for their solution exaflops supercomputers. The approach is demonstrated at the examples of topical problems of material science, condense matter and dense plasma physics where atomistic modeling is necessary to apply. The correspondence is established for each problem between phenomena studied and computational cores number needed. Modeling parallel programs scalability is shown as well as perspective of the modeling methods predictive ability extension with the increase of computational cores number and / or use of special architecture (graphical processing units). The following problems are considered: 1) surface modification at processing of metals by sub-picosecond laser pulses, 2) radiation-induced aging of nuclear reactors fuels, 3) phase transition kinetics in metastable liquids, 4) methane and hydrogen gas hydrates structures and computation of their properties, 5) polymers multiscale models, 6) dusty plasmas, 7)ion recombination in liquid and gaseous dielectric media at discharge break and relaxation, 8) electric double layer between graphite and electrolyte, influence of electron-hole electrode structure on capacity. Predictive modeling reliability is checked by comparisons with experiments. The modeling methods hierarchy, which is necessary to describe properties of matter at different space and time scales, is considered in frames of the multiscale approach. Density functional theory (quantum molecular dynamics) is applied at the deepest nm/pm scale to model electron dynamics and to construct effective interaction potentials between particles. Classical molecular dynamics modeling is used to treat moving atoms systems up to micro-scale. Kinetic theory and continuum mechanics is used to proceed with micro-scale. Particular attention is paid to the exchange of information between different scales, i.e. to the unified description of systems from nano to micro levels. Parallelization efficiency comparison is performed for three classes of problems at fat tree and torus topologies (in Russian).

Key words: atomistic modeling, electronic structure, molecular dynamics, multiscale modelling, radiation aging, laser ablation, nucleation, hydrates, polymers, dusty plasma, electrochemistry, parallel efficiency.

article citation

http://psta.psiras.ru/read/psta2014_1_191-244.pdf

 

• Содержание выпуска

 

Adress: Ailamazyan Program Systems Institute of the Russian Academy of Sciences, PSTA Online Journal, 4 a Peter the First Street,
Veskovo village, Pereslavl area, Yaroslavl region, 152021 Russia
Phone: +7-4852-695-228.       E-mail: info@psta.psiras.ru.      Website: http://psta.psiras.ru

© Electronic Scientific Journal "Program Systems: Theory and Applications" 2010-2017
© Ailamazyan Program System Institute of RAS 2010-2018