Exploração da linguagem Rust para o desenvolvimento de um path tracer paralelo

Abstract

Computational generation of images is a technology in use for a multitude of applications. Traditionally, languages such as C++ have been used to write systems for these kinds tasks. Rust, a new systems programming language, has as one of its goals to also fit this role. This work had as its goal to re-implement the SmallVCM renderer, originally written in C++, using Rust, to evaluate if this new language has the potential to be a viable alternative for these tasks. Parts of the program have been translated, including one of the available rendering algorithms, trying to maintain a similar program structure, resulting in a renderer with equivalent functionality, capable of using this algorithm to render images equivalent to the original. Performance profiling of the new program shows that it is able to generate these image in less time than the original program compiled with a similar optimization back-end, even whithout having received any algorithm changes.