Struggle to see where a Nvidia competitor would come from, analyst says

Well, I mean, there's not enough superlatives to describe the performance. That's the problem. And I think that, you know, everyone keeps looking at the expectations. There was a lot of nervousness ahead of these results, but they smashed it again, demand is kind of, you know, outpacing supply. So you have to look at it and go, there's a lot of potential here still. And as we saw, as you say, the the re pivoting around kind of these AI factories, I think as Jensen like to call it, and some of the new opportunities he's seeing with AI kind of sovereign investments as well. It's an astonishing time for this company. When it comes to what we're seeing on the buying side, Alphabet, Microsoft, Google, Amazon counting for about 45% of the company's data centre revenue. We know that they're buying these chips for training purposes. We know there's inferencing around the corner as well. But what happens once these machines are trained up? What does the demand actually look like for these high end chips? Well, I think it seems at the moment that there's plenty of appetite to keep investing in fitting out those data centres. We think that the models are going to get more complex. They're going to need more and more processing power. So it sits well for NVIDIA from that perspective. But they are clearly looking to open up other opportunities as well. So we heard in the earnings talk of sovereign AI, the fact that, you know, government investments like Switzerland, for example, are looking to train up their own models. And I think we'll see more of that as people start to make strategic bets on AI at a national level as well. When we're looking around and thinking like who can compete with NVIDIA at this point? We know companies like Microsoft and others are looking at their own semiconductors for data centres, but actually it still looks at this point like NVIDIA is the only game in town. However, we are seeing AMD and others come to the market. Does NVIDIA, as you look over, let's say the next year or so face any real test when it comes to to a competitor? I think it's hard to see where that comes from right now. You know, they've locked out a lot of the supply of the core GPU technology. They're very embedded in terms of, you know, the developers are very wedded to the platforms and and and interfaces that they provide. So I struggle to see where that competition comes from. But no company in the semiconductor space can afford to ignore this opportunity. So of course they're all investing and they'd all like to take a slice of Nvidia's pie. And you mentioned that that sweet spot Nvidia's in now, large demand supply still slightly constrained. Yeah. Do you see that working itself out at any point where that demand and supply sort of match up? Well, I think Jensen said in the earnings he doesn't see any issues in terms of, you know, demand pushing right on into next year. We've got this transition on the platform from the Hopper to the Blackwell chips. That's going to be interesting whether some of the customers pull back and wait for the new chips. But then they revealed as well that the the new chips will start shipping in the coming quarter, which I think was again ahead of people's expectations. I want to pick up on the Sovereign AI theme because what we've seen from Microsoft Copilot PCs, so effectively a on device this week is that the AI gets very personal, it's very close, it's very invasive in some ways. It's watching everything you do. And after all the data privacy concerns we've seen through the lens of social media, it's telling us, well, we don't want that to play out badly on an enterprise way or even on a personal level as we get into an AI universe. It does tell you that those data centres where the data is stored and how it's used very close to to where it's going to be used. So it means it's a different cycle, doesn't it? This time round that you're going to see countries having to step up on these investments. You're saying enterprises going to have to step up. It's different, isn't it? Well, there's so many facets to AI. So obviously there is that there are countries that want to have their own assets. They want to have it, you know, on their own territories. They also want to learn about the technology. They see it as a big catalyst for future technology investment. But then we're also seeing a lot of discussion around the on device AI. So, so that's not necessarily related to what NVIDIA are doing, but some of the other players in the market and some of those models which are being developed that sit on the device. And therefore, in theory, all of your data remains with you and doesn't go up into the cloud. And that's going to be a very, very hot debate over the next few quarters.