Font Size: 

The technique for semi-automatic parallelization of loop operators is proposed. It is based on loop tiling and data serialization and uses rewriting rules to transform programs. The technique allows to extend GPU capabilities to deal with big data volumes that outfit internal GPU memory capacity. It can be applied to utilize clusters consisting of several GPUs. Applicability criterion is specified and a semi-automatic proof-of-concept software tool is implemented. The results of the experiment demonstrating the feasibility of the proposed technique are given.