Seminar

Abstract
The development of shaders is probably one of the most eye-catching topics in the domain of computer graphics, as shaders are the key to unlock the power of GPU to make the final output on screen look stunning. Although GPU-based shader is a recent phenomenon, the idea of shading could be traced back to 1977, after the release of a low-budget film called Star Wars IV: A New Hope. Lots of new concepts and algorithms were introduced in 1980s and graphics APIs such as OpenGL were released in 1990s, with the aim of adding more flexibility to shading capability, but all of these designs became possible by the early of 2000s, when graphics hardware had become more sophisticated and fast enough.

Two main categories with different focuses on images output have emerged after that: offline rendering and online/real-time rendering. The former one concentrates on the quality of images and has been widely applied in film and advertising industry for decades; while the latter one prioritizes the promptness of output speed, which has been accepted as a default solution of the game industry.

Due to the continuous upgrade of graphics hardware, the gap of image quality that laid between offline and real-time renderers is becoming less and less visible, especially after the appearance of modern game engines such as Unity and Unreal Engine which have democratized high-end development skills even academic concepts to ordinary individuals and independent studios.

In this 10-minute presentation, the challenge that game-ready shaders need to face nowadays will be explained and several potential solutions are presented based on researches on Unity engine and its graphics-rendering pipeline. Techniques related to real-time ray tracing will be shortly covered as well, as a potential game changer and possible trend in the following decades.