SUMMARY
The article deals with the issue on using hardware in teaching distributed programming. It is established that the integrated use of software and hardware in the study of distributed programming is one of the important methods for the development of professional competencies of software engineers. It was found that the effective distributed programming teaching requires a large number of resources to develop these calculations, which are compact enough to fit into one course and sufficiently autonomous. One of the methods of studying distributed computing is the use of a training computing cluster. The development of a training computing cluster includes the following main stages: development of training tasks that require the use of a computing cluster; hardware development; installation of the operating system; designing a data transmission network between cluster nodes; configuration of the main node and cluster management programs; installation of software for the development of distributed computing. Therefore, the use of the cluster will help the development of professional competencies of future software engineers in the study of the discipline Parallel and Distributed Computing, which is taught to software engineers in the specialty of computer studies. Computers and Raspberry single-board computers were used for comparison as computing cluster nodes. After comparative analysis, the initial computing cluster was modeled using Raspberry single-board computers. Testing has shown that on large amounts of data, for mathematical computing, the Raspberry single-board training computing computer makes computing one and a half times more efficient than a personal computer. The analysis of development of professional competences at modeling, development and use of initial computer cluster is carried out.