Wrapping up and summarizing Nutanix Enterprise AI
The Nutanix presentation at AI Infrastructure Field Day focused on enterprise AI solutions, emphasizing giving customers a solid technical understanding of Nutanix Enterprise AI (NAI) and its role in addressing key customer challenges. The discussion highlighted the curated model catalog, offering pre-configured and customizable models, and the ability to incorporate new cutting-edge models, even within air-gapped environments, easily. This approach provides control over models and data, which is particularly relevant for customers seeking sovereign AI solutions and needing to deploy AI models in their own environments.
Nutanix also emphasized the “deploy once, inference many” model, allowing for the creation of a shared service model where multiple applications can connect to deployed models via endpoints. Furthermore, the session touched upon the simplification of sizing, as NAI streamlines the deployment of models, making the process straightforward. The speaker reiterated the benefits of NAI as an application running on Kubernetes, offering flexibility and portability. The presentation concluded by discussing the future of distributed inference across multiple nodes, acknowledging its importance and status as a planned future development.
A key takeaway from the presentation was the growing demand for sovereign AI, driven by geopolitical factors and specific terms of service that restrict the use of certain models in certain regions. Nutanix recognizes and actively helps its customers address this need by providing the necessary tools and infrastructure to enable control over AI models and data within their own environments. The company’s commitment to adapting and evolving its AI solutions to meet the rapid advancements in the AI landscape was underscored, ensuring that Nutanix remains a relevant player in the enterprise AI space.
Presented by Mike Barmonde, Sr. Product Marketing Manager, Nutanix. Recorded live in Santa Clara, California, on April 24, 2025, as part of AI Infrastructure Field Day. Watch the entire presentation at https://techfieldday.com/appearance/nutanix-presents-at-ai-infrastructure-field-day-2/ or https://techfieldday.com/event/aiifd2/ for more information.
Transcript
So let's quickly summarize. Okay. What we talked about today, we talked about this curated model catalog, which is absolutely cool, right?
We give you a lot of the different models that we prescribed, that we've configured. It's guided, it's automated, it's secure, it creates these endpoints, very great stuff from a model type, even your own custom model. And you have that freedom to experiment.
You have the ability to take cutting edge models like, you know, in this case a deep seeker, whatever comes next tomorrow or whenever, and quickly upload it, including an air gapped environment. And that means control over models and data in these type of environments. You have air gapped environments, you have all these different places.
The freedom of choice becomes really critical. I will say, to add on to what I heard earlier, someone mentioned customers, not at liberty to say all those different stats, but I will say this, many, many of our customers that are choosing an AI are creating and wanting sovereign AI control of their AI in their own environment. And they're willing to spend the money to do that.
The current geopolitical situation and what comes next or whatnot, is creating this necessity. And we are right in the middle of helping them out and getting feedback on that, which is cool. Deploy once, inference many, right?
Create this shared service model. So we talked about the capability and Jesse mentioned as well as slash weenie about taking a bunch of these models, deploying 'em once, and then actually connecting multiple different applications via these endpoints. And then sizing made simple, right?
Jesse talked about the nuances and how complex sizing can be when it comes to understanding the infrastructure. And it does a lot of that hard work making deployments pretty much a snap when it comes to deploying these models. Okay, questions, comments?
Clap. Clap plat again. Run any CNCF.
Kubernetes. Kubernetes. Kubernetes.
Kubernetes N AI is an application, a set of containers that runs on any on Kubernetes, wherever Kubernetes is. N AI can be pretty cool. Number two, remember, deploy all the models you want, any of the types that you like, we help with that.
You can do the pre-configure ones or you can bring your own, you can also then create a secure endpoint. Again, I think someone mentioned earlier, what is an endpoint? It is an API that's connected to an LLM.
This little bundle you give to your developer, it's open AI compliant, they're ready to roll, and that means people can actually test this before it goes out as well. And that includes developers and application people for ai. Make it super simple.
Reiterate, reiterate, reiterate. Alright, We saw it. We saw it again.
You told us what you told us. One quick question. Yes, Thank you.
Well done. When I'm looking at any model and I wanna run LAMA four and it doesn't fit on one node and I need to run it across multiple nodes, can I inference across multiple nodes at this point? Or is that something that might come in the future?
Ashwini. It is A future. Um, okay.
And uh, and thanks for asking that. That's kind of reiterating the ask for us, but um, it is, it is something that we planned for future. Um, and I I can give you a little bit of reasoning why it's, it, it's a future versus already, Although that's really hard.
Uh, yes, that too. Of course the Community Is still figuring it out at the end of the day. Yes.
Yeah. Yeah. Um, I, I think one of the, uh, pieces was, you know, four or 5 billion model from meta was one of those forcing functions where they're like, oh, you absolutely need to run this on two nodes.
3, which is 70 billion, which was as good as the four or 5 billion. I know. Dodge that One.
And then people like, oh, we don't need that anymore. Right? So that's, that's essentially why it's been going a little bit back forth.
Yeah. Then LAMA four, now it's back. Yes, exactly.
It's one of Those only that if you look at the terms of service, guess where you can't use LAMA four in Europe, right? So now you have geopolitical and geographical blocks between these things about how to run. So again, the sovereign AI piece becomes very critical based on their terms of service.
Wow. Is that A-G-D-P-R thing? The European or just I, I, I think I, I don't, I can't infer about what it means, but Oh, okay.
I I definitely think it has to do with the way that their AI regulation and I see the, the way they do things is set up. So yes. Okay.
Possibly. Got it. Yeah.
