원문 보기
Listen, did you see the post from David Ellison this morning, Jim? The company is now called Skydance. There will be a lot of all kinds of changes. New studio executives are emerging at Warner. " Digger" comes out this weekend. These Disney layoffs and challenges. This entire segment is in such a difficult situation right now.
It takes a lot of courage to intervene. Wow, look at this. You know, usually we see the Paramount logo on the mountain, but... Wow. This is what the future looks like. There were 12 shares, they went through all this, and now there are nine. I'm trying to understand how it all... Remember how Faber buried them?
I joked that David "buried" them because they had too much debt. They need to do something. Although I have to say, I'll give them credit for that. Do you remember when Zaslav came to Warner Brothers Discovery? Where were they, seven of them? He said, "Listen, this is a $28 stock." I had dinner with him, my wife, and Pam.
He tells my wife— you know, she hates dinners with business people, but she likes Pam. We are not on air. And he said, "You know, I told your husband that a stock at $7 would go up to $28." And the wife replied: “Oh , really? Yes? How is that? How is that? Then he turns to Pam and says, "You know, don't you?" I asked, " Seriously?" And he said: "Oh, it will be 28." And I...he was right.
This is a good example, because the lending conditions are a little different now if you have a lot of debt, Jim. How much should we reduce now compared to those times? It's very risky when you do something like this, it's like ... Yeah, listen, it's like a house. You have to pay your mortgage, and the interest rates are just outrageously high right now.
Nice to see...What about you and Chris, who came to the position of co-CEO at a bad time? Everyone says he knows how to cut costs when needed. I know people are disappointed with the stock performance, but I will still say that he saved the company. The balance was in terrible shape. This is not enough.
Everyone who bought them in 2018 when he was on the show. He's absolutely right to feel terrible about it, but there was some value in Elon's actions, and I don't want to dismiss that. Yes, I mean, this industry has been through a lot between COVID and tariffs. True, but you know what people look at, they say: look at this deal, look at what Chris Cox did.
He, he did a lot of different things that, uh ... Well, just the whole intellectual property strategy. I mean movies, whatever. Yes. No, I mean, they see the problem, you know, my father sold toys for 3 years, and it was even worse than when he sold carpets, but a little better than when he sold duct tape, but not as bad as when he sold corrugated cardboard .
This is a series of mediocre products. So, he was selling food bags to fill your restaurants because people are so terribly stingy. They take leftover food home. This is my father. They come home, perhaps with suitcases. Yes. Oh no, he had this wonderful package with a tie on top, made of heavy, sturdy cardboard.
Because he didn't want to be eco...he was eco-conscious long before anyone else . I, I remember because he loved these so-called packages. He really loved them. He loved this job. He would have liked it today. That's right. That's right. I would really like it . Campuses that are measured not in square feet, but in hundreds of megawatts, and eventually, soon, gigawatts.
On a broader level, this is n't just a story about GPUs anymore, is it? Data centers attract an entire ecosystem around them, right? This includes compute, storage, networking, as well as cooling, power, and more. And the restrictions are also changing. There is definitely demand and capital. We have known this for some time, but we also know that on the side of limitations there is the connection to the electricity grid, we also see the problem with water, and we must not forget about community acceptance.
All of this really affects the speed of development. And we see that this is really turning into a massive construction of industrial infrastructure. And for investors, this expands both opportunities and risks. Huge capital expenditures—that's something you're also watching, Ron. Tell me more about it, what you think about it.
Communities are resisting to some extent, but these data centers are getting bigger and bigger, are n't they? Yes, I think the momentum has already been set and it will continue. In Lisa's opinion, I believe this will lead to 5 gigawatts of capacity by the end of this year, and this is largely fueled by hyperscalers' capital spending on data centers.
By the end of this year, this amount is likely to reach $750 billion. So, we're actually looking at what's driving this process. Well, part of it is that continuing to train advanced models will require high-density AI infrastructure. Additionally, model inference is becoming more widespread, and this will continue to use data center resources, as well as be more widely distributed at the edge.
So, this also provides more opportunities to expand AI functions across the ecosystem. It is also important to add that regarding the US data center market, we expect an average annual growth rate of 12% over the next 3 years until 2030. And I believe that this coincides with Goldman Sachs' estimates.
I think the reason for this is that, despite some difficulties and concerns of local communities, there are many compensating factors. We already see that the construction of data centers affects, for example, election cycles, so I believe that this situation will continue to develop. Growth is so inevitable that even with certain , I would say minor, resistance factors and other constraints, it will not stop the overall momentum.
Yes, we may see some slowdown due to community resistance or something else, but you both seem to be on the same page. It will see this inherent growth that continues here. Um , when we think about specific names, Lisa , what comes to mind in terms of leadership and market share? I've already mentioned a few of them , it's not just a graphics processor race anymore .
AMD and Western Digital. AMD makes the situation much more interesting, as it is no longer just a story about alternative GPU number two. I think AMD's data center business reached almost 7 billion last quarter. And over the last year, we've seen Meta, OpenAI, Anthropic—all of them together have committed to using up to, I think, 14 gigawatts of AMD GPUs.
We know that energy efficiency has also become part of the economics of this process. But I think customers are interested in how much AI can we get out of every watt, out of every rack? And I don't think AMD needs to win to replace Nvidia. They simply need a larger share of the rapidly growing market.
As for Western Digital, AI is n't just about computing, right? We are talking about the data economy. And of course, this is only going to continue, and so data storage may be a somewhat underrated component of the ecosystem. Um, the training that Ron was talking about, the data output— they're the focus, right?
They require a lot of computing resources. But as we get deeper into output agents, there's going to be a kind of data flywheel happening, right? Data is created constantly . And so we're not just creating new data. I think what will happen is that historical corporate data will be brought back because of its value.
And this, in my opinion, is a tailwind for Western Digital. Um, the company is gaining more and more exposure to the demand for hyperscale cloud storage for AI . So, you know, the more data we have , the more it has to be stored somewhere, and that's where Western Digital comes in. And we're thinking about what we need with AI, and Open AI's annual revenue , which turned out to be about 20 billion less than expected, hit a lot of companies, including Western Dig.
That was something that was partly included in this story. And Western Dig, which rose to almost $800. I discovered this because I was reviewing some of these titles as they had already seen drops this week. You looked at these names here. Ron, this is the AMD story you were talking about regarding Western Digital gaining market share and some concerns about that.
You know , will we slow down or grow? I vote for growth and I think in addition to what Lisa has already aptly shared, there is the fact that AMD, for example, acquired World Nvidia is approaching a milestone that seemed $6 trillion. Meanwhile, stocks are just below record highs after five straight winning sessions.
And according to Raymond James analyst Sim Leopold, Nvidia may still have significant room for growth. His price target is $515, one of the highest on Wall Street. But the bigger story isn't jus spend huge amounts. large-scale investments in AI infrastructure. And even though the cost of borrowing remains high, .
And that's it , that's the the development of an entire industry. Nvidia is right at the epicenter of this. There is also a great deal of confus will be only one winner and everyone else will lose. But the reality can re specialized cloud providers, and there are many directions simultaneously. Nvidia positions itself as one of the critical infrastructure providers that connects many of these pieces.
But there is one problem. The bottleneck now is not just companies' desire to ge pushed memory prices up, and memory isn't the only problem. Energy becomes enormous amounts of electricity. They need networking, cooling, storage, and modern components, and it all needs to work together. So while investors are focused on demand for GPUs, the next stag those processors.
focused on computation. Then memory became an important part of history. Now the industry is increasingly focusing on fast information transfer between all components so that expensive equipment does not stand still. Think about what this means. You can build huge computing power. You can install a lot of memory, but if those systems can't interact effectively, you'ward more complex reasoning and agent workloads, the architecture becomes even more complicated. Different types of computing devices must constantly interact with each other.
댓글 0