Russian New University (RosNOU) (Russian: Российский новый университет, abbreviated РосНОУ) is a Russian non-profit institution of higher education.
The article discusses which hardware accelerators are advisable to use if you need to train large language models on your own servers, rather than in the cloud. The main types of accelerators are analyzed — GPU, ASIC, FPGA and others. The cost of ownership of different solutions for local use when installing 256 accelerators is compared separately. As a result, it is concluded that the choice of equipment should take into account not only the speed of calculations, but also the costs of electricity, cooling and software compatibility.
Subject. Limitations of classical approaches and specialized methods of estimating the value of football clubs. Objectives. To identify and systematize the limitations of classical approaches and specialized methods of estimating the value of football clubs, as well as to assess the discrepancy between the results obtained by specialized methods. Methods. At the first stage, the analysis of academic literature and industry sources was carried out. At the second stage, a comparative analysis of the cost estimates published by the analytical platforms Forbes, Football Benchmark and Sportico for 2025 was carried out. Using the example of Manchester United, Borussia Dortmund and Juventus football clubs, available specialized methods (Markham, Solntsev and Kinnaird Valuation Model models) were tested using data from annual reports, Transfermarkt and other open sources. The results are compared with the estimates of the analytical platforms. Results. Classical approaches have limitations that make them largely inapplicable without significant modification. Specialized techniques also exhibit significant methodological shortcomings and produce varying results. The average deviation between the estimates of Forbes, Football Benchmark and Sportico was 22%, and the maximum was 48%. The spread of estimates between the Markham, Solntsev, and Kinnaird Valuation Models ranges from 32 to 220%. There are also discrepancies between the estimates obtained using specialized models and the estimates of the analytical platforms Forbes, Football Benchmark, and Sportico. Conclusions. Each of the existing methods of estimating the value of football clubs has significant limitations. Specialized techniques only partially take into account industry specifics and retain their own shortcomings. The results of the study confirm the need for further improvement of the methodology for estimating the value of football clubs. The findings can be applied by analysts and investors when conducting a cost analysis of football clubs for a more informed choice and interpretation of the evaluation results.
This article examines the possibility of mitigating the tort liability of a member of a collegial body in the presence of subsequent good-faith actions by the person who committed the civil violation aimed at preventing further harm to the property interests of the debtor and its creditors. The advantages and disadvantages of introducing a legal mechanism that would allow for such behavior to be taken into account as a circumstance that, at a minimum, allows for a reduction in the amount of civil liability are examined. A conclusion is reached on the need for a balanced approach: the introduction of a mechanism for taking into account good faith behavior after unlawful actions requires clear procedural criteria and clarifications from higher judicial authorities in order to avoid abuses and maintain the predictability of legal regulation.
Relevance . In the context of the digitalization of the public sphere, digital media are becoming a key agent of youth political socialization, transforming the traditional system of socializing institutions. Algorithmic mechanisms of social platforms shape new patterns of information consumption and political participation, generating cognitive and socio-political risks that require scholarly interpretation. The purpos e of the study is to identify the mechanisms through which digital media influence youth political socialization and to substantiate the need for a transition to a comprehensive model of state governance of digital socialization within the framework of youth, information, and educational policy. The research objectives include analyzing the transformation of agents of political socialization in the digital media environment, identifying the mechanisms of digital media influence on youth political consciousness, and developing a typology of risks associated with digital socialization along with their governance implications. Methodology . The study is based on systemic and institutional approaches, the theory of political socialization, and the concept of the network society. The methodological toolkit includes comparative analysis, generalization of statistical data (Mediascope, FOM), structural-functional analysis of digital influence mechanisms, and risk typologization. Results . It is established that digital media transform the vertical model of political socialization into a horizontal, algorithmically governed system in which digital platforms and influencers act as key agents. Three interrelated mechanisms of influence: cognitive, emotional, and behavioral are identified. Individual, social, and political risks of digital socialization are determined, and their governance dimension is revealed. A structural contradiction is substantiated between the formally established responsibility of the state for shaping youth political culture and the de facto dominance of algorithmic systems of private platforms. Conclusion . The study substantiates the need to shift from reactive regulation of the digital environment to a proactive platform-based model of public governance, which implies cross-sectoral coordination of youth, information, and educational policies, as well as the institutionalization of media literacy as a strategic element in fostering a stable political consciousness among youth.
We calculated the angular distribution of ions backscattered from a solid target. The results of previous theories are generalized for the case of arbitrary angles of ion incidence. Approximate analytical theory is developed also.