Fascinating read! The final approach of interpreting the shaders initially, while compiling them in the background for greater performance, sounds very similar

Fascinating read! The final approach of interpreting the shaders initially, while compiling them in the background for greater performance, sounds very similar to what just-in-time compilers do. If you think about it, the problems they face are also kind of similar: both systems are confronted with unpredictable, dynamic "source code", and both want to achieve both high performance while avoiding the lag introduced by ahead-of-time compilation, so it makes sense that a similar solution approach might work.