Render farm for highly realistic images in a beowulf cluster using distributed programming techniques

Now-a-days, photorealistic images are demanded for the realization of scientific models, so we use rendering tools that convert three-dimensional models into highly realistic images. The problem of generating photorealistic images occurs when the three-dimensional model becomes larger and more complex, so the time to generate an image is much greater due to the limitations of hardware resources, about this problem is implemented the render farm, which consists in a set of computers interconnected by a high-speed network that provides a strip of the global image distributed in each participating computers with the intention of reducing the processing time of highly complex computational images. The research was implemented in a high-performance Beowulf group of the Universidad de Ciencias y Humanidades using a total of 18 computers. To demonstrate the efficiency of a rendering farm implementation, scalability tests were performed using a 360° equirrectangular model with a total of 67 million pixels, the work is carried out to achieve highly complex renderings in less time to benefit the direction of the research.