M. Zibul

Along with the powerful computational tools of modern information processing systems are characterized by three important characteristics: storage capacity, speed of information access and secure storage duration. For applications that work in these areas, especially databases and storage systems, the factor limiting performance is most often input-output subsystem. Data growth, increased storage requirements for reliability and speed of access to the data resulted in the allocation of the storage subsystem to a separate computer system. The role and importance of such systems is determined by the ever-increasing value of information in today's society, access to data and management are essential for the execution of business processes. The key requirements for modern data storage systems are: reliability and fault tolerance, data availability, the availability of funds management and control, performance, scalability. Information security is an increasingly important factor affecting the architecture of storage systems, taking into account the introduction of an increasing number of new applications in public and private spheres. There is a continuing trend in the implementation of data storage technology to protect data from unauthorized access. Encryption is an effective means of addressing data protection. Many firms manufacturers of storage systems implement encryption of information in their devices and systems. With the growth of digital content meeting the increasing demand for archives requires a robust platform that provides flexible scalability, high availability and performance for faster retrieval of data required in an environment of open systems and mainframe environments. When you create a local storage device emulation and access, as well as modern data storage systems for the implementation of the most important features of modern storage systems, namely the encryption / decryption of data and search data, it is proposed to use the graphics processor unit (GPU). The results presented show that despite the limitations of architecture GPU (lack of dynamic memory allocation, the inability to synchronize between the executable blocks of shared memory bank conflicts, particularly access to global memory, etc.), their use in the field of data storage systems provides significant benefits.