Microsoft CANCELS Major AI Data Center Deal… (NOT GOOD)

[00:00:00] Speaker 1: Microsoft just canceled a major AI debtor center project in the UK because the power grid reportedly won't be ready until the 2030s. And this is all happening at a time when NeoCloud stocks are getting hit with many skeptics and bears arguing that, hey, they're going to fall off a cliff because demand is one of the big concerns. Now, I've covered the concerns around CoreWeave's debt, the broader NeoCloud sell-off, and investors questioning whether these companies can actually deliver on their enormous AI infrastructure commitments. And now, just a few days later, we have Microsoft backing out with a project getting canceled because of the grid. And this is where I think things get interesting because on the very same day, reports emerged that the company behind this project, Nscale, was negotiating yet another $45 billion deal with Anthropik. So, Microsoft is walking away from one project because the electricity isn't there in the UK, while Anthropik is potentially committing another $45 billion to the same company in the same day. Now, obviously, as a Nevis shareholder, I think that there's a bullish argument to be made here that people might be overlooking, which I want to cover in this video. But there's also a ton of risks as well that I also want to cover in this video. Without further ado, let's jump right in. Okay, so let's talk first and foremost about what happened at Nscale. But first, if you're not subscribed to this channel yet, then I urge you subscribe. It helps the channel grow, and it helps me make more videos of this nature. Let me know if you enjoy this type of content and if you want me to be talking about it more often. Nscale is a UK-based AI infrastructure company that develops data centers and provides compute capacity to customers like Anthropik as well as Microsoft. Last year, they announced plans to develop a major AI data center in Essex, England, to support Microsoft's Azure business. This facility was expected to house approximately 23,000 NVIDIA GPUs with operations beginning in 2027, and the local power grid reportedly cannot provide enough electricity until 2030s. And so as a result of this, Microsoft has decided to cancel its agreement for that specific site. Now, they haven't canceled their entire relationship with Nscale. I mean, Nscale even said that their relationship is good with Microsoft, and both companies have essentially made this distinction clear that they're still working together. But it is still a very big setback because Nscale is going to IPO. And this company was reportedly seeking a valuation around 30 to 35 billion dollars for their IPO. But now, as a result of this news, investors are actually pushing down that valuation to about 16 to 19 billion, meaning that Nscale has lost about half or close to half around their IPO value. And this isn't the only execution challenge that this company has had. Their Norway facility has experienced delays with just 6.4 megawatts of operating capacity out of 30 planned megawatts, according to a recent report. And this also raises an important question out there around how much weight are investors putting into these enormous data center pipelines before the actual infrastructure is running. So there was this post on X circulating that compared Nscale to IREN arguing that investors should be placing more value on companies with a demonstrated operating capacity rather than projects that remain largely on paper. And I think this is a very legitimate concern because execution is a big distinguishing factor. And I think that's something that has so far set Nevis apart and will continue to do so, right? We shouldn't assume that the same companies are going to have the same execution challenges. And this is one of the things that I saw on X. It's a fallacy to say that because Nscale can't deliver, or because the UK can't deliver specifically, that Nebius is automatically going to get hurt as a result of this. And I'm going to make both the bull and the bear cases here. But I want to start off by saying that there is an execution layer here. And if there's a company that's not held back on their timelines, and if there's a company that's almost always held back on their timelines, well, the question then is, what are the unique factors about that scenario? And is it, in fact, execution? But then we got another major development, because Nscale is reportedly negotiating currently a second computing agreement with Anthropic with an approximate worth of $45 billion. And if this second deal gets finalized, well, we're potentially talking about almost $90 billion in commitments between these companies. So there's incredible demand out there. But there's a couple of caveats. So first of all, the agreement still is under negotiation. And second of all, Nscale still needs enormous amounts of financing and infrastructure to deliver on these commitments. So it's not as clear cut and dry. In fact, its IPO prospectus for this company disclosed that as of the filing, it hadn't secured binding financing for the infrastructure required under its original Anthropic Agreement, let alone its new Anthropic Agreement that's apparently under negotiation. And this is, I think, one of the pieces that makes this story very interesting. We're seeing customers commit extraordinary amounts of money on AI compute, while the suppliers on the other side, and the infrastructure is massively lagging behind the financing and the physical infrastructure is one of the big concerns, because they're bottlenecks of what's needed to actually service that massive demand. And so let's bring this back to Nebius, right? I want to ask the question, is this bullish or bearish for Nebius? And I want to actually present both cases, starting off with a bearish argument, because that's more straightforward. And I think that's the conclusion that a lot of X has jumped to on the back of this Nscale news. And that is simply to say, well, if Nscale cannot deliver a major Microsoft contract because of electricity constraints, what is stopping something similar from happening to Nebius? And I've spent a lot of time discussing this on this channel previously, whether it's live streams or the recorded videos, between the power, the construction timelines, the financing, and the infrastructure availability, these are all becoming significant bottlenecks for the entire industry. And the market is increasingly concerned about these risks. This is why you see some of the pullback in the price action markets are wanting to take risk off the table, just in case these other factors are not coming through. If a project is getting delayed, if that revenue is getting pushed out, if the financing and the development costs continue to build in the meanwhile, if the debt is coming due and you're not building a data center any faster, well, all of these things are why the market is pricing in the risk with the emphasis on execution right now. Nscale doesn't help this fact. And look, Nebius is not immune to these problems as well, because at the end of the day, Nebius has an enormous development pipeline and it's going to need significant capital to be able to deliver on that pipeline. So the Nscale situation, in my opinion, is something that we should pay attention to. And I think it's sort of like the canary in the coal mine as a warning for the rest of the industry that execution is a very big part. So that's the bearish argument, okay? Let me now talk about the bullish argument, because this is where I think the opposite argument becomes very compelling for Nebius. What if Microsoft's cancellation had nothing to do with the demand profile, right? What if Microsoft still needs that computing capacity, but Nscale simply can't deliver it? Well, there's a couple of different situations. And this is where Nebius, I think, has a potential advantage. This is where my mind went to first. We know that Microsoft is already a Nebius customer. They've already established a commercial relationship with Nebius, with a company that has been delivering Microsoft capacity already and has a track record of doing so consistently since that relationship started. In its last earnings report, for example, Nebius confirmed that it delivered every Microsoft capacity tranche up to that point. This is important. They haven't faltered. They haven't delayed. They haven't canceled anything. They've delivered everything on time. And look, personally, I think Microsoft is going to expand their relationship with Nebius regardless of this Nscale news because they just need to. Microsoft's AI infrastructure requirements are not suddenly going to go away just because Nscale can't deliver on them. And I don't think the existing contract that Microsoft has with Nebius is the extent of how much their business are going to do together. They still have a lot of computing needs. But when you have another supplier running into delays, it potentially creates an additional opportunity for Microsoft to expedite their renegotiation or their expansion with Nebius. Like, let's say you're Microsoft, you need the compute capacity. You have these deliverables and Nscale all of a sudden fell through. They say, we can't give you anything until 2030. You're going to look to your other suppliers and you're going to say, okay, who actually can give me that that I'm working with? And Nebius comes top of mind because they haven't failed on any timeline so far. And obviously, you're more likely to go with an existing provider than go with a new provider altogether and establish a new relationship. Now, I'm not speculating that, you know, Microsoft is going to take the Essex contract and hand it to Nebius. I'm not saying that because the relationship is still good with Nscale. And we have no announcement that that's happening right now. But what I am saying is that expanding an existing relationship with Nebius is a realistic possibility. And I think that possibility was definitely on the cards by default. But this Nscale news, I think has a possibility to even expedite that expansion. And, you know, another thing to consider here is pricing right Nebius management has already discussed keeping some future capacity available, rather than locking it up in long term contracts. And this is very interesting now, because look, in 2027, they're charging way higher for this. And they're saying, we're deliberately holding back this capacity from selling it because we think we can sell it in the future. Now, if computing capacity remains scarce, which it's seeming like it is, and demand keeps growing, which is seeming like it is, that could make available capacity from providers like Nebius even more valuable, especially to large hyperscaler existing contracts like Microsoft. Now, obviously, Microsoft would run a long term contract to be locked in. But let's face it, the reality is that they're probably not going to get that because Nebius has no interest in doing that when they can sign shorter term deals for higher prices overall. And this is, I think, a distinction that matters, ultimately why I wanted to make this video. It's not about nscale. It is about the reliability in the execution costs of these neoclouds for their ability to deliver on existing capacity. And those who can are going to benefit disproportionately than their competitors who are getting hurt because of these macro constraints, because of the infrastructure, because of the power and whatnot. Now, there's another few developments that I want to highlight before wrapping up, because right now, it's an interesting time. Obviously, we have Nebius earnings around the corner. And I think that's going to be particularly interesting. Given the current market sentiment, I think management is going to address that sentiment. I've talked about my expectation, how I think Nebius is going to raise their seven to nine billion dollar exit ARR guidance. That's still my prediction. I think they're going to raise that this quarter. But I also want to hear more about the 2027 CapEx plans and how Nebius plans to finance this expansion and whether capacity deployments remain on schedule, because as of last quarter, they were, but we want to hear an update. Right now, I think the market more than anything needs reassurance that all of this investment can eventually translate into real revenue. And I think if Nebius delivers strong results, raises their guidance, or announces another major customer agreement like a Microsoft expansion, for example, at a much higher price, I think that could go a long way towards rebuilding the confidence in the stock and sending it above 200 to 250. I got asked in a live stream yesterday, can Nebius go back to 250? And my answer was, look, if they announce good news, and if they reiterate that guidance, and they raise it, then absolutely they can. But we need company specific catalysts to reassure the market right now. Because right now the macro is dominating. There's a lot of noise out there. Like even yesterday, Elise AI, which is a South Korean AI infrastructure company, they fell sharply after debuting their IPO. And there was a lot of discussion on X around, you know, this AI data center company falling because the demand wasn't there. And I think that's more reflective of poor sentiment than it is reflective of poor AI demand. Because a weak IPO, literally around the world in Korea doesn't tell us whether Nebius is or is not executing on its contracts. But it does give us an insight into how skeptical investors are becoming towards the data center sector in general. Nebius is participating at GDC Berlin. It's going on today as I'm recording this on the 20th through the 22nd. And I think Arkadi is also speaking. And we've already seen announcements at its European AI builder initiative, as well as Nvidia related training collaboration. So I would expect in this conference, we might get some Nebius news as well, or at least some positive developments overall from Nvidia around the whole space. That's my expectation. It's not really, you know, a confirmed announcement we're going to see. I think the keynote is tomorrow, which by the time you're watching this, it's probably already happened. But overarchingly, nobody can deny that the sentiment is negative right now. But even still, incremental evidence of stronger partnerships and execution can send the stock higher, can reassure the market of Nebius' confidence in buying that stock. So when I look at N-Scale and when I'm asked about what I think N-Scale is doing, I don't see clear evidence that the demand for AI compute is collapsing. What I do see is I see an industry where the demand is overwhelming. It's, it is enormous, but at the same time, infrastructure is becoming increasingly difficult operationally. And as a result, this is both an opportunity and the risk for a company like Nebius, because ultimately the risk is that these same constraints on the macro also serve as constraints for Nebius. But the benefit is that companies that can execute and companies that can stay on track with their deliverables are going to have repeat business in a bigger way, because all of these other companies, they can't deliver, right? Microsoft still needs that compute by 2027, as was the initial agreement. By 2030 might be too late. So the question is, if you're Microsoft right now, what do you do? Do you go to somebody new in the market or you go to somebody that you've been working with, that's proven to be reliable, that's delivered everything on time? And this is exactly what I'm predicting. I think Microsoft is going to expand their contract with Nebius because they already have that relationship and because they have that overwhelming demand and because Nebius has proven that they can execute. I'm still bullish, but ultimately more and more, I think in the eyes of the market, it will come down to execution because announcing capacity is very easy, but actually delivering on that capacity we're finding is increasingly difficult. Thank you so much for watching. I hope you enjoyed this video. Let me know your thoughts in the comments below. Do you think Microsoft is expanding their contract with Nebius? And if so, when? I'll let you know if anything is announced at GTC Berlin. We're going to stay tuned over the next couple of days, but that's all going to be in another video. Thank you so much for watching. Subscribe to the channel for more. I really appreciate it. And I will see you very shortly.