Collaboration with “Dubna” University

*Tutorial “Practical introduction to MPI technology on HybriLIT cluster” is published (download).

International IT-School “Data Science”

International IT-School “Data Science” of Dubna State University is a University organizational structure which provides IT-specialists’ training for the development of megascience project computing (NICA, PIC, LHC, FAIR, SKA, etc.), Data Science, digital economy, and other perspective directions.

The creation of International IT-School is a joint initiative between the Joint Institute for Nuclear Research (JINR) and Dubna State University taking into account the development of unique world-class scientific projects at JINR. Educational programs of International IT-School are based on personnel requirements of JINR and other high-technology economic organizations; moreover, they are realized with their help.

Bachelor’s and master’s theses

2019, 2018, 2017, 2016

2019

  • Kakenov Meyrzhan Beybutkhanovich
    03.04.02. – Nuclear and Elementary Particle Physics, Nuclear Physics
    A study of the dibaryon model of nuclear forces in light nuclei with A = 6
    Master’s Work
    Supervisers: E.V. Zemlyanaya, V.I. Kukulin
    Scientific adviser: V.N. Pomerantsev
    ABSTRACT
    The main goal of the work is to study the new concept of nucleon-nucleon interactions based on the dibaryon model of nuclear forces as applied to the description of light nuclei with A = 6, within the cluster model of light nuclei. A software implementation of a variational solution of the Hill-Wheeler integral equation was developed to find the parameters and wave functions of light nuclei with A = 6, carried out on the HybriLIT heterogeneous cluster of the Multifunctional Information and Computing Complex (MICC) of the JINR Information Technologies Laboratory. The results of calculations within various models of nuclear forces with the available experimental data are analyzed.
  • Chernyshev Yaroslav Mikhailovich
    01.04.02 – Applied Mathematics and Computer Science, ISAM
    Personal account of the heterogeneous platform “HybriLIT” user
    Master’s Work
    Superviser: Dr. Streltsova O.I.
    Scientific advisers: Butenko Y.A., Belyakov D.V.
    ABSTRACT
    The paper is devoted to the development of the project of the personal account system of the user of the heterogeneous platform «HybriLIT». Based on the results of the review and the analysis of existing web services (personal account system’s),the web service was designed and developed for running, editing and viewing the statistics of running tasks, as well as providing remote access to the user’s task management system and ensuring the correct work of a lot of users simultaneously. The result of the work was the project of the user’s personal account of the heterogeneous platform “HybriLIT”.

2018

  • Erofeeva Kseniya Sergeevna
    27.04.03 – System analysis and management
    Development of a complex of programs for studying the system of long Josephson junctions on hybrid computing architectures
    Master’s Work
    Superviser: Dr. Streltsova O.I.
    Scientific adviser: Bashashin M.V.

2017

  • Butenko Yuri,
    Federal State Budgetary Educational Institution of Higher Education Dubna State University
    The problem of gamma-neutron separation for the DEMON spectrometer using machine learning methods
    Supervisor: Prof., Korenkov V.V.
    Scientific advisers:PD. Dr. Streltsov A.I., Dr. Streltsova O.I.
    The goal of this paper is to research the applicability of machine learning methods to process and classify experimental data obtained from the DEMON detector and to perform a comparative analysis of the methods’ effectiveness as well. DEMON type detectors are one of the basic measuring elements for ATLAS and ALICE experiments usually used to detect neutrons and gamma quanta.
    The conventional software package for this detector allows to map all events into energy diagrams that are recorded and saved for later statistical and mathematical analysis. The main problem of statistical processing of the obtained data is to separate and identify which events correspond to neutrons or -quanta.
    The double integration method used in the conventional framework to solve the problem does not enable clear identification and separation of these events at certain energy intervals, that is a motivation to apply machine learning methods to this problem.
  • Savelyeva Ekaterina
    Software development for mobile neutron spectrometer
    Mastrer’s thesis, University of Dubna, ISAM
    Superviser: Dr. Streltsova O.I.
  • Kabiev Musa
    Estimation of microscopic optical potential of elastic scattering 6He + 12C in parallel mode 
    Bachelor’s thesis, University of Dubna, department of nuclear physics; Eurasian National University (Kazakhstan)
    Supervisor: Dr. Streltsova O.I.
    Scientific advisers: Dr. Zemlyanaya E. V.
  • Daribaev Ayzat
    Simulation of  DEMON neutron spectrometer efficiency calculation
    Bachelor’s thesis, University of Dubna, department of nuclear physics; Al-Farabi Kazakh National University
    Supervisor: Kozulin E. M.
    Scientific advisers: Dr. Streltsova O.I.
  • Mayorov A. V.
    Development and deployment of the web-service for experimental data analysis  on HPC platforms using ROOT package
    Master’ thesis, University of Dubna, ISAM
    Supervisor: Korenkov V. V.
    Scientific advisers: Dr. Streltsova O.I., Dr. Vala M.

 

2016

  • Ushakov S.
    Parallel algorithms for the tasks of multidimensional integration
    Bachelor’s thesis
    Superviser: Dr. Streltsova O.I.
  • Mayorov Aleksander
    Development of  a module for statistics gathering using web-sites for user activity monitoring
    Bachelor’s thesis
    Supervisor: Mikheev M.A.
  • Mareev Ivan Romanovich
    Investigation of the effectiveness of parallel algorithms for calculating the value of integrals using the GPU in the OpenCL language
    Bachelor thesis
    Supervisor: Ayriyan A.S.

Summer schools for students

2019

Trainings for 2nd and 3d grade students of the Dubna University on Parallel Programming Technologies is held on the basis of the heterogeneous computing cluster HybriLIT.
The key aim is to learn parallel programming technologies that allow developing applications for carrying out computations using modern computing architectures.

Educational program on HPC technologies based on the heterogeneous cluster HybriLIT (LIT JINR)

2019

  • O.I. Streltsova, D.V. Podgainy HybriLIT heterogeneous computing platform: ecosystem for parallel computing, application development, ML / DL and data analysis.”