Abstract: The role of Large Language Models (LLM) in new generation integrated development environments (IDEs). Tools such as GitHub Copilot, IntelliCode and Alice Code Assistant are explored in the context of their use in programming. The authors examine how LLMs enable the automation of key development tasks, including code autocompletion, error detection, refactoring, and code generation, which result in increased development efficiency and improved code quality. Special emphasis is placed on how LLMs affect developers' cognitive processes, such as problem-solving abilities, creativity, and professional skills. A review of existing integrated development environments that utilize large language models. LLM functionality for code autocompletion, fragment generation, error detection and correction was evaluated. Comparative methods were applied to evaluate the effectiveness of LLM compared to traditional development tools. Special attention was paid to analyzing the cognitive load caused by the use of LLMs and assessing their impact on the creative process. The novelty of the research consists in the complex analysis of LLM application in modern IDEs, as well as in revealing their potential for increasing developers' productivity and improving the quality of program code. It is concluded that LLM integration into IDEs allows not only speeding up the process of code creation, but also considerably increasing its quality due to intellectual support and automation of the routine tasks. However, while the benefits of integrating LLMs into IDEs are clear, limitations related to cognitive load, ethical issues, data security, and the need to maintain a balance between automation and development of programmers' skills are also identified.

Keywords: development process optimization, machine learning, software systems, code completion, artificial intelligence, code improvement, programming automation, Integrated Development Environments, Large Language Models, data analysis

Abstract: The subject of this study is neural networks, trained using transfer learning methods tailored to the specific characteristics of the dataset. The object of the study is machine learning methods used for solving speaker verification tasks. The aim of the research is to improve the efficiency of neural networks in the task of speaker verification. In this work, three datasets in different languages were prepared for the fine-tuning process: English, Russian, and Chinese. Additionally, an experimental study was conducted using modern pre-trained models ResNetSE34L and ResNetSE34V2, aimed at enhancing the efficiency of neural networks in text-independent speaker verification. The research methodology includes assessing the effectiveness of fine-tuning neural networks to the characteristics of the dataset in the speaker verification task, based on the equal error rate (EER) of Type I and Type II errors. A series of experiments were also conducted, during which parameters were varied, and layer freezing techniques were applied. The maximum reduction in the equal error rate (EER) when using the English dataset was achieved by adjusting the number of epochs and the learning rate, reducing the error by 50%. Similar parameter adjustments with the Russian dataset reduced the error by 63.64%. When fine-tuning with the Chinese dataset, the lowest error rate was achieved in the experiment that involved freezing the fully connected layer, modifying the learning rate, and changing the optimizer—resulting in a 16.04% error reduction. The obtained results can be used in the design and development of speaker verification systems and for educational purposes. It was also concluded that transfer learning is effective for fine-tuning neural networks to the specific characteristics of a dataset, as a significant reduction in EER was achieved in the majority of experiments, indicating improved speaker recognition accuracy.

Keywords: deep learning, neural networks, speech processing, feature extraction, speaker recognition, speaker verification, dataset, fine-tuning, transfer learning, pattern recognition

Abstract: The object of the research of the article is the need and the possibility of developing the resources of an academic environment, that is, on a non-commercial basis, specialized graphic editors that provide a visual design environment for instances of applied mathematical tasks of various types. Software available on the market, as well as their extensions, approximately corresponding to the specified task, are considered. It is shown for what reasons these tools turned out to be unsuitable for those applied tasks or those customers with whom the authors had to deal earlier. The research method is based on a comparison of the functionality of the software available on the market and their extensions, as well as the author's reserves, with the applied tasks on which the authors emphasize. The novelty of the study and the conclusions made is the proposal to separate the task of visual design and automatic generation of computational code, as is customary in existing products. It makes sense to transfer the creation of a computing code to an applied mathematician working with a government customer, but at the same time organizing an effective division of labor between the customer and the mathematician. At this angle, the tasks of the graphic editor are simplified. On the other hand, such an editor can be made more flexible.

Keywords: MATLAB, mathematical graph, user interface, visual design, web browser, mathematical calculations, graphical editor, Simulink, JavaScript, JSON

Abstract: The article reviews a system of management of robotic system in which a person acts as an executant. Here both person- executant and robot- executant are equal agents of the heterogeneous complex. It is supposed that agents of the heterogeneous complex regardless of the characteristics have a common goal and are able to adjust their actions during the execution of the task. The authors describe the structure of interactions within the heterogeneous complex. The article provides information on the implemented system of interaction with human as with agent of mobile robotic system. The authors describe structure of human-computer interaction in the heterogeneous robotic system. The heterogeneous robotic system is considered as a multi-agent system. The novelty of the approach described by the authors in the article is in considering the person as an equal agent of multi-agent system equal to intelligent robots. The prototype of the system of interaction with a person as an agent of robotic system was developed and tested in the faculty robotics laboratory of MEPhI. At present researchers are focused on improving the quality of interaction and unification of interaction system.

Keywords: control system, robotics, multi-agent systems, human-computer interaction, voice messages, heterogeneous systems, mobile robots, intelligent systems, robotic systems, speech recognition

Abstract: The article describes main principles of the remote direct memory access technology (RDMA) in highperformance computing systems. The specific attention is given to the algorithms for data transmission to and from network. The author reviews standard algorithm for data transmission and RDMA algorithm on the example of TCP/IP-based networks. The author studies the most popular computational architectures supporting remote direct memory access, proves the effectiveness of RDMA for high-performance network architectures on the example of 10 Gigabit Ethernet. Using the theory of sets and calculations of processor\’s work cycles during the processing of network instructions and data the author compares performance of the nodes of computing system transmitting data via standard algorithm and nodes that implement RDMA technology. The article explains the expedience of RDMA implementation for compute nodes of high-performance architectures. In modern computer architectures computing nodes are interconnected via network, so a considerable amount of data is transmitted through network. One of the methods of improving the efficiency of computing nodes and the network as a whole is to offload processor of network computations with the remote direct memory access technology (RDMA). RDMA-mode is a technology for transmitting network data without CPU, directly from the network adapter buffer to application buffers, thereby freeing the CPU from processing network data. In high-performance computing architectures, systems the amount of incoming and outgoing network traffic is very high and the processing of network data occupies a large amount of CPU time, so the proposed technology of remote direct memory access is highly relevant to offloading processor of a specific computing node and hence to improving the performance of the whole system . Compared with the standard data transmission in which the processor of a compute node system completely handles network data RDMA technology frees the CPU from network load and makes it possible to focus all production capacity to internal tasks increasing the speed and reliability of the whole system.

Keywords: high-performance architectures, remote access, direct access, network technologies, processor performance, CPU performance, RDMA, network adapter, TCP/IP, data transmission, GigabitEthernet

Abstract: In this article the author shows that the process of building a system of technological process automation requires creating a hie rar chy of interconnected dia grams: struc tural – functional – an electronic circuit dia gram – wiring diagram. The author des cribes basic components of the intelligent CAD «Controlics». The article shows basic principles of a new method of automatic generation of functional circuits, based on a compact representation of a set of measuring and executive circuits in the form of trees of circuit solutions and the allocation of some branches of the tree into special templates from which to derive functional diagrams. The author presents an algorithm that allows to identify and to exclude from consideration the schemes that are technically unrealizable on the level of the electrical connections. A formulation of the problem of generalization of functional circuits structures based on rough sets theory and stages of its solution is given. As a result the author deÞ nes production rules, the application of which allows generation of control systems circuits in automatic mode. The article describes the methodology, software and technological features of the proposed method and functional features required for the implementation of the software.

Keywords: Software, design, scheme, connection, chain, pattern, verification, generic, heuristics, products