Earnings Call - NVIDIA CORP

Welcome back to the main stage. We are talking about, for the next two sessions, AI. AI from going from ambition to implementation. What does that take? For this session, you will have a keynote, followed by a high-level roundtable. They will go back to back. Your moderator, coming up a little bit later, is Michael Lake from Leading Cities. But your keynote, who is coming up right now, is Paolo Guglielmini, who is the VP for EMEA at NVIDIA. He is expecting the room to fill. Gentlemen, do come forward, slowly walking, finding a seat, take a seat. Fantastic. As we delve into AI and the practical possibilities for what it can do for our urban cities. Carlo, Guglielmini, good morning, ciao, welcome. The stage is yours. Thank you very much. Good morning. Thank you. Such a beautiful energy, such a beautiful smile. Welcome, everyone, and thank you for having us. It's a pleasure, it's a privilege to be here on the main stage and get us kicked off for the day. When you look back at the organizations that you represent, most of the great cities have been built on great infrastructure, right? Could be related to transportation, could be rail, could be roads, could water, could power. But the paradigm is going to shift for the next 10 years, right. Building great infrastructure is going to mean building great intelligence. Intelligence that's available to your citizens and residents to create a better experience and to create opportunity, but also intelligence that is going trigger and kick off that economic flywheel that you need for the place in which you live and you want to live. So for the next 10 minutes or so, I'm going to try and focus on how AI factories are going to be a booster and a foundational element in the way you look at building intelligent cities for tomorrow. And we definitely want to start by celebrating some of you in the audience and in this incredible conference that have already gone very far along the road of building very intelligent cities. As you know, your organizations, your cities, municipalities and regions have never had so much data available to you. If you think through water utilities, pollution, traffic, battery chargings, think about mobile devices, think of GPS, drones, satellite data. There's so much available to get the finger on the pulse of what's going on in your cities today. But the gap between the data that's available and the data that we use for decision-making has never been this great. So the logos that you see on screen are innovators and leaders that have started on the journey of closing that gap, of using AI to be able to draw insight, contextualize their data and ultimately deliver a better experience to your citizens and ourselves. So when we look across the board of Europe and Africa and Middle East, where a lot of this work is already ongoing, but there's so much more that we need to do, knowing that technology is moving incredibly fast, right? At NVIDIA, a lot of what we do is to try and keep our own finger on the pulse of technology. Great connectivity with frontier labs, great connectivity with natives. What we do ist to try to understand how the art of the possible changes every single day, right? And when you look back at the last three months, how eventful those three months have they been? So we constantly need to reevaluate what is feasible, what can be scaled, what is realistic, what is economically feasible. When we look across the board of our region, we see three areas of applications that are so transformational and are becoming more and more widespread. First of all is digital twins for sake of planning. Secondly, it's about vision AI that's implemented for sake off improving operations. And thirdly, it is about agentic AI that is used to deliver citizen services and resident services. So I want to double click on a few of these examples. So the first one comes from France, right? This is the surroundings of the Dijon area in between central and east of France where the department door has worked with local technology developers to go and build perfect digital mock-ups of their assets of their buildings and we're talking about more than 500 buildings right coming from different eras and these are highly intelligent digital mockups this is not about purely geometry right these are mock-ups that are enriched with data, maintenance data, records, utilities records. The reason why we want to do that is to be able to plan for those buildings and assets as we go forward, right? We want to have an intelligent record. We want to be to run better operations on those buildings and valuable assets today, right, with all of the advantages that mining instantly rich information can give us, but also we want to be able to do city planning into the future. Simulation has never been so powerful, right? One of the privileges of being with NVIDIA is that we get to look across this field of AI adoption, across financial sectors and healthcare. We look at public sector, we look at defense, we Look at the work that we do in automotive, in robotics, right simulation and multi-dynamics simulation is one of the fastest growing areas of adoption so simulating cities and urban planning into the future has never become so affordable so quick and so valuable and of course the likes of akira and think deep have been building on models that come from nvidia and are running this simulation in france on sovereign cloud as we speak the second example comes from the world of vision agents so once you have built perfect mock-ups of your buildings well now you want to integrate those buildings in the reality of a city that's alive right you want To integrate those Buildings in the rest of your infrastructure the city is pumping with events and opportunities and challenges and real-time sensor data so what you want to go and build is a true city brain, right? And this is what Palermo has done so effectively over the last months. When you think of Palermo, you probably know how challenging in terms of traffic a city that has such an ancient history, a maritime city can be. I come from Genoa in the north of Italy. The layout of Genoa is very, very similar. And the constant challenge How do we modernize and how do we promptly intervene when something goes wrong, right? So Palermo went off and built a center of excellence that is incredibly capable to fuse together sensor data, data that comes from a multiplicity of facets, and really shorten the time to emergency response, really make traffic more effective, more fluid, work on safety, and work all the way to waste management. And also in this case, what they did is to create the right ecosystems for themselves. They grouped system integrators together with infrastructure experts in the likes of Dell, and they've worked with local ISPs, K2K in this moment, to go and build that context and also develop those applications on top. And again, here, you probably know NVIDIA for our infrastructure. You probably know Nvidia because of the ecosystems that we try and pull together. But NVIDDIA is one of the richest libraries, if not the richest library, of open technologies, software, and toolkits that you can go and consume to kickstart your adoption of AI. Last but not least, we want to talk about agents for citizen services. And I find this example mind-blowing at the scale and the speed at which it came about. And this is something that you can reproduce in your organizations as we speak. And the third example takes us back here to Barcelona. In Barcelona, an AI native called One Million Bot has created an agentic platform. This agentic platforms utilizes a model called Alia. Alia has been developed on Barcelona infrastructure at the Supercomputing Center in collaboration with the government. And Alia is a model that is highly specialized, highly sovereign for Spanish language and for all the dialects and the slangs that are important to you. So it could be Galician, could be Basque, could mean anything that's relevant to address the largest possible population of your citizens. So what 1 Million Bot did is to create a suite of agents that are able to book appointments and help you through procedures that are going to be helpful in terms of urban planning, giving you access to all of those offices and services that very often are hard to reach in the first place. It's going to give you access to information. It's gonna manage service inquiries. We all know how much of our livelihood and experience as citizens is driven by the level of digitization of the cities that we live within. And here too, the city is in control. It's in control of the applications. It's en control of data. It's control of permitting and permissions of who gets to see that data. Because we're dealing with highly sensitive pieces of information, of course. Let's see a little more about the art of the possible that you see flowing on screen. As you look across these applications from agent, from civil engineering, two things to keep in mind. One is control and one is scale. All of the organizations that we have talked about were incredibly concerned by the span of control over data. technology has moved so fast in terms of sandboxing agents, in terms of sovereign infrastructure, in turns of giving you control of the data that you utilize to create those services. Secondly, scale. If you've noticed in these three examples, people went off and built an infrastructure foundation. They have created the right ecosystem of competencies And as we know, skills and competencies very often are the number one constraint in order to get started, even more than capital. But then when that foundation comes into place, you build for yourself an opportunity to be really creative about the art of the possible and the use cases. You build for yourself a foundation to go and proliferate use cases, right? Very, very often the organizations you are within have a lot of legacy. We have data that is nested. We have silos. We have applications that have been built over the course of tens of years. A Gentic AI is perfect to build that layer on top that harnesses the power of data. So as we talked about at the very beginning, infrastructure build out for the next ten years is going to mean building intelligence that is at your fingertips to help you run your cities more effectively, more economically, and go and kickstart that economic flywheel. So once you've gotten inspired in terms of use cases, how do you get started? How to think about kicking the ball and getting the ball rolling? Two things. One is focus on the building blocks of building AI. And the second thing is building the ecosystem. What do we mean with building blocks of AI? Well, this is a famous metaphor that I think Jensen donated to the world and I think is quite effective. It's a simple way to think about building AI. AI is a five-layers cake that starts with energy, continues with compute and infrastructure, needs models to run on, and ultimately value creation happens at the level of applications. Now, think of this in the context of your organizations. A lot of the cities and municipalities that we work with have a treasure in the form of powered shell and power infrastructure that is just waiting to become a source of tokens and intelligence for your communities. When it comes to models, the world in the last six months has come a very long way in terms of creating tokens in a very economical way, giving you control. The future is always going to be multi-model. The future is becoming increasingly large models and highly specialized small models, frontier and closed working hand-in-hand with open technologies. But when I'm looking at my city and where I come from and your cities, the economic opportunity will come from applications. I have joined NVIDIA from a business that operates in robotics and software and physical AI six months ago. If I have to mention one takeaway of my experience in the last six months, it's an undeniable truth. You go across regions, economic opportunity will happen where compute resides. Talent will stay where compute infrastructure is built. The lion's share of the startups that are created today are AI-native startups. And entrepreneurs and innovators will always follow where computing infrastructure is built. This means that if you get the five layers right, economic opportunity in the form of innovation and startups and spin-offs will come from the top end of those building blocks. and last but not least it's about building an ecosystem it's about building a thriving ecosystem of competencies and technologies and brains that come together to build and make your use cases real what does that mean it means AI infrastructure it means cloud it means telecommunication providers for the next years the name of the game will be consumption of tokens in a way that is highly decentralized that reaches applications more and more at the edge in the physical world and certainly it's about solution providers that can help you build a more holistic view of these competences where all of these ecosystems come together in your context is very often a center of excellence right more and More we see cities right that appoint chiefs of AI and center of excellence that are mandated to bring all of these competencies together. And we want to be there with you in your centers of excellence to be able to point you to the right competence for the right journey that you're on. Also, what is important is that when we talk about AI build out, very often it's about gigawatts, right? And it's above market caps, and it's below staggering numbers of investments but your ai journey can start from something that is very small can start from something there is not inexpensive but certainly portable and incredibly powerful i am staggered by the number of very famous ai natives that operate in europe today that have started from something that is as simple as a DGX Spark. So when you look at your AI journey, having the right level of ambition is fundamental, but getting started with a use case that is really transformational for your communities and think about it in modular terms is fundamental to get going. I have seen in my technology career so far, I have seeing a lot of technology shifts, right both in the physical world and in the digital world when it comes to ai i've never seen a technology that offers an opportunity for return on investment from pocs as quick as with this technology shift so i want to leave you with three simple thoughts and we're going to keep the conversation going in the context of the panel focus on the use cases that are the most transformational for where you are and the means that you have at your disposal. Bring together the right ecosystem, and then look at your territory and the resources that you have available to identify opportunities to build compute, and then grow from there, okay? Thank you very much for your attention. I want to invite on stage Mike Lake, and we're going to get growing discussing more about the out of the possible for AI factories changing the intelligent cities of tomorrow. Thank you very much.