The parallel capability of modern GPUs with openGL

Hey y’all!

Deepening my understanding of Vulkan and the capability modern GPU have to parallelize the threads of shadowing programs, I found this magnificent article collection (Breaking Down). I’d like to know if any information exists regarding the barriers points of parallel capabilities in openGL has by default, the ability to parallelize threads. I assume that things like using the same VAO or VBO on the same subroutines are part of the barriers by e.g. In the end, I understand that it’s the access to linked memory the thing that determines that openGL by default, parallelizes the workload threads

There’s an overview wiki page regarding memory access and barriers here: Memory Model
but for details, check the spec: OpenGL 4.6 Spec.