Good question. Let's start with the TAM. Last quarter, we talked about a TAM of around $8 billion. So I think that it's clearly -- the reality is proving that the number is significantly higher. The market has still very, very different numbers. I said we have seen numbers of the 10 times the $8 billion or more than 10 times the $8 billion. It's still unclear. We have to, I think, a little bit more. But clearly, it's a market that is expanding. Of the 30 gigas that we talked about, as of September 30, only 20% of it, one small portion were in our pipeline. The rest were contracts that we started to -- with customers since then. And then today, if you ask me today this morning, roughly half of the 30 gig are in pipeline, the other half we're working on it. And what we're looking is -- will they happen in the next 2 years, where do we see our product is suitable to do what they want. And generally, I think we are fine. So what's a big change from telling you a quarter ago, this is an $8 billion market requiring this very, very complex capabilities to today. I think there's a big change in terms of what we can do for what our technology and Fluence in particular, can do for data centers. And I would say the way to think about it is that there are 3 needs. One is what we call interconnection flexibility, the ability to manage your -- the energy demand in a way that you can interconnect easier to the grid and you can manage and the distribution companies or the service provider can manage your demand to keep the -- so that is by itself, I would say, today, the biggest driver. People who want to connect quickly to the grid and want to ensure that the data center meets the availability of the grid and can give the assurances to the grid operator that they will not disrupt the grid. And we can do that today. That's what work. This is -- there's no -- we have no need to improvements in our technology stack to be able to do it. So great. The second one that is also in a rising need or rising need is backup power. Historically, we haven't played that game. But with our costs coming down as they are and our ability to -- our density improvements, we can now provide backup power and significantly reduce. I won't say eliminate, but significantly reduce the need for diesel generators. So that's the second need that we're seeing. We can accelerate interconnection to the grid, and we can reduce some of the cost of the diesel generators by providing backup power. The third one is the one we have talked about in our last call, this power quality, this idea that we can -- we will have to manage the variability of energy demand by AI data centers. That -- if you ask me today, that hasn't been -- the first thing is that there are other technologies that can address that. That's the first one. The second one is that it is a need that is not as big as we thought it was going to be. So it's probably around that $8 billion number. And it is something that data centers, when they look at what they're doing, their speed to power is a much more important element than this one because the other one they can manage in some other way. We are committed to delivering the 3 products. The interconnection flexibility to accelerate interconnection, the backup power capabilities. And these two that we can do today, and we're very well positioned to do. Smartstack is the densest project in the world. It is a project that because of the [indiscernible], the way we are designed provides very good safety, better than, I would say, very, very good. And then third, our cybersecurity, our total control on software, our ability to ensure that no one else can get it. So the power quality is something we're working on with our inverter manufacturers. We'll get it resolved quickly, but it's still a work in progress. But we thought that was going to be a gating item the backup power is going to be a gating item for us to serve this market. That's no longer the case. I would say it's a cherry on the top. If you can deliver the last 2 and this one is great, but it's not a gating item. So great market, multiples of what we told you in terms of what we do, and we don't need to do a major technology. And my last point, we can -- we don't have a clear view today. This is just starting on how much we can capture. What I will say, we are very well positioned to do safety, density. Some of our competitors are claiming density, which is 20% to 25% less than what we can do. So that tells you we can do very, very well, and we are -- we have -- we hired Jeff. Jeff comes with knowing how to serve this market. He's been one of the structure, go and get this done. And this is not only happening in the U.S. This is a global phenomenon. We have in our pipeline. It's mostly U.S. today, but we're starting to see pipeline coming both out of Australia and Europe. So sorry for the long answer, but that we're excited about this opportunity.