Antonio Neri

Think about hospitals where decisions needs to be put in the hand of the doctors using AI models with MRI machines and all the type of machinery.

to provide better care.

And look, there is a lot of robotics surgery taking place these days.

Experience is at the edge in the hospitalities, right?

So a lot of data is generated, but instant gratification.

So that's why a lot of the inference will be done in distributed environments, but fundamentally is a hybrid model where both training inference will live and coexist in centralized environment and in a distributed environment at the edge.

That's right.

And that's why the networking is the core fabric by which we enable the deployment of AI.

Absolutely.

And at HP Discover, we announced a number of sovereign solutions

so that the sovereign AI clouds can be deployed faster.

Certifications, you know, regulations that are built inside this GreenLake cloud allows you to deploy, whether it's an air gap or northern gap, with a peace of mind that you are in compliance with those regulations, which is a big effort because to do that yourself takes so much work and so much effort.

And so we build it at the core inside the fabric of the solution itself.

But I have to say there is 10 to 12 customers around the globe, countries, that have the capabilities to deploy this at scale, and the conversations are accelerating, and the build-out of data center is also accelerating.

having everything under their control, because including the impact to their culture, to their identity.

Countries can lose their identity very quickly by using a model that was trained somewhere else other than their country.

While you can live with that, the fact that you are now giving data and context to that model is an opportunity to lose that identity.

So there is a geopolitical factor involved here.

And so people realize this is a big change and they want to have it under control.

That's exactly, yeah.