My guess would be tight software integration between Nvidia Arm CPUs and Nvidia GPUs that would not be available to other Arm vendors or users of Nvidia GPUs with Intel CPUs.
CUDA provides software lock-in to Nvidia GPUs. This would provide software lock in to the combined Nvidia Arm / Nvidia GPU platform.
CUDA provides software lock-in to Nvidia GPUs. This would provide software lock in to the combined Nvidia Arm / Nvidia GPU platform.