they announced a cloud service NIM that hosts LLMs using Nvidia software stack. https://developer.nvidia.com/blog/nvidia-nim-offers-optimized-inference-microse

they announced a cloud service NIM that hosts LLMs using Nvidia software stack. https://developer.nvidia.com/blog/nvidia-nim-offers-optimized-inference-microservices-for-deploying-ai-models-at-scale/ https://developer.nvidia.com/blog/nvidia-nim-offers-optimize... so there is a push towards a platform company, but probably not well explained in this article.