Visualization and image processing tools integra-tion is a tendency of recent years. 
Image restoration is a topical research trend for methods and tools de-velopment 
of degradations and noise compensa-tions that are added to the recorded data. 
The effec-tive evaluation of image restoration algorithms needs state-of-the-art computer platforms. 
It is shown that using NVIDIA graphic accelerator for general purpose problem (GPGPU) 
computing with central processing unit (CPU) speeds up image res-toration evaluation 
in 4 times in comparison with only CPU realization. The common recommenda-tions that 
are based on finding and results of differ-ent image processing realizations on GPGPU-CPU platform 
for computations accelerations are given in the end of the article.