Optimization of resource-intensive computing tasks on graphics processor

Produced analyze of the possibility of running computational tasks on graphics processors (GPU), investigated why such efficiency is achieved and what measures need to be taken to achieve it. Demonstrated the advantage of computing on graphics processor units in general purpose of the example matrix, which was achieved by optimizing the code to run in parallel on the GPU. Offered a software library that provides API to perform the tasks of processing data arrays optimized for execution on the GPU.