Compiling OpenCL code using clang for AMD platform
In this video we are going to see how you can compile OpenCL code using clang, the main goal is not to get executable code but to get the readable assembly to investigate what our code is doing. In a future video, we are going to see how that code will actually execute.
We are gonna compile for AMD platform, currently running on a Ryzen 1700 OC to 4.0 ghz and water cooled, the card is a rx 570 4 GB. I did not test this yet but most likely you don't even need to be on AMD platform clang should be able to cross compile nicely.
In the future, we are gonna use this technique to compare what differences we have between different kernels.
In case you are interested, I teach courses about Maya API, optimization and how to write GPU accelerated deformers for Maya at: www.riggingdojo.com
As usual, if you liked the video like, share and subscribe, it really helps the channel out!