[qayxc]

NVIDIA used CUDA to establish industry-wide vendor lock-in on GPU compute.

Google uses TPUs to try and establish an industry-wide vendor lock-in on deep learning.

Same old, same old.