저자: JimCramer · 외국인
요약 보기
엔비디아는 AI 인프라 수요 확대로 시총 6조 달러에 근접했고 Raymond James는 목표가 515달러를 제시했다. 병목은 GPU가 아니라 메모리·전력·네트워크로 옮겨가고 있으며, 에이전트 AI 확산이 데이터센터 투자와 네트워크 장비(시스코 등) 수요를 더 끌어올릴 것이라는 논지다. 테슬라 3분기 인도량은 486,532대로 컨센서스(약 457,000대)를 상회했다는 보도가 소개됐다.
다듬은 글
Speaking of Musk, we got Tesla supply figures this morning. On this subject, let's turn to Dom Chu . Hello, Dom. Good morning, Karl. Good morning, Jim. So, Tesla's third-quarter delivery figures now stand at 486,532 units for the period from July to September. That beat analysts' forecasts of about 457,000.
Tesla said it ended the second quarter with its largest order book since 2023. So, the exact number is 486,532. That's better than last quarter's deliveries of just over 480,000, and it beat consensus estimates from both Tesla itself and FactSet. We see that the stock is now up a little over 2%, and again, the consensus analyst opinion may be changing.
To give you some context, Carl and Jim, the current FactSet consensus price target is $376, implying a 6% upside from current levels, with only about 44% of analysts recommending a buy and 46% a hold. Here's how it all looks against the backdrop of the third quarter supply data, Carl. I'll pass the word back to you.
Well, look, it's nice to see that when gas prices are going up so much, people are thinking that maybe it's more profitable to switch to an electric car. I also think— remember, I said you buy Tesla because I think SpaceX is a great company. By the way, SpaceX, I should have mentioned when I quoted those "colossi", it's SpaceX that borrows and charges me for computing power.
Yes, they have more computing power. So you have SpaceX, which I think could show explosive earnings growth, and it would be a great decision to buy Tesla when that happens. This is a good indicator for Tesla. This is still, I mean, essentially what BYD did in one month. I know. Yeah, I mean , I was looking at the BYD last week in Italy, and you know, it looks like a $60,000 car.
Well, what can I say? Obviously, the suicidal nature of Europe is simply incredible. Yes. And of course, yesterday we got the data from GM, Jim, where electric vehicle sales seem to have halved. I mean, these are difficult comparisons before the tax breaks expire. Yes, I understand, but look, you look at the price of Ford, you look at Stellantis, you look at GM.
For me, questioning any numbers that GM publishes is wrong. They actually...Mary Barra doesn't do self-promotion. And I'm promoting Mary Barra. She is outstanding. Let's look at the situation in pre-market trading . Obviously, there are a lot of conflicting factors this morning as we analyze the macroeconomic situation and these, to put it mildly, disappointing employment figures .
Although the story with the Fed, after the speeches of Waller, Jefferson, Bowman, and now and after that, gives the stock market an opportunity to breathe a sigh of relief. Certainly. Squawk on the Street will be back in a minute. The additions amounted to $8.2 billion. And I expect that this acquisition alone, simply put, will make the company smarter.
That is, they will be able to better understand the requirements for spatial intelligence and develop advanced AI models simply to expand the sales of their portfolio. And this, of course , applies to both their GPU portfolio and, of course, the CPU segment. And I think that's going to be important in understanding the big picture in the aerospace and defense industry, and why they're already forecasting that they're going to achieve significant performance over the next few quarters.
And I think in addition to that they brought in ZT Systems. And I think that will be important in terms of getting the intellectual and design capabilities again needed to ensure that their portfolio adapts to the needs of advanced AI models that are constantly changing. And specifically, regarding Western Digital, I don't think this is a unique case just for them.
This actually applies to the entire memory space as a whole. And this is certainly the segment that will grow rapidly. I see Western Digital's share at around 12%, depending on how you value the market. But if you look at Samsung, SK Hynix, and also Micron , all of these players, I think, will continue to show very impressive financial results.
And this includes, for example, achieving a margin of 80%. So, I think , yes, the failure with OpenAI, and I'll characterize it that way , will have, you know, a very, I would say, short-term impact. But if you look at the overall long-term picture, it will be growth and even more growth, if you don't take into account some meta-events.
So, I think this will be very good news for the entire ecosystem. I'm also thinking about agent-based AI. I was just reviewing a report on Palantir from Goldman Sachs. They actually spoke very positively about Palantir. But getting back to the big picture of AI agents and what might happen there, Lisa, you're also looking at that part of the story, right?
We look at data centers and think about some of these big names that we've already mentioned. I mean, Nvidia has been taking off lately with all the deals it's making with different companies. I mean, even today we see it in the news. So, what do you think about agent-based AI, data inference, and all that?
As I said before, model training gets all the headlines, right? But if we look at this emergence, and not even an emergence, but the rapid growth of agent-based AI. This is one of the main topics at every conference I go to. So as companies move more towards data and agent-driven, as I said, it all boils down to data.
How much data can we create, how much can the physical infrastructure handle? Because a lot of companies, and I hear this across industries—Ron, you probably do too— are banking on agents, treating them like interns or employees, to make sure they're identifying the exact workloads that we need to automate; Which ones will help empower people?
Agent AI is taking center stage, and everyone I talk to is getting involved in this process. Yeah, and your final overall view, Ron, I'd like to know what you're projecting now and maybe for the coming years. Certainly. I believe that agent technologies will become an integral part of overall workflows in organizations, and this is becoming quite evident at the corporate level.
Yes, many are still hesitating and watching to see how soon we can entrust agents with these tasks with built-in safeguards, and I think that will be a key point. In fact, we see the ecosystem benefiting from the increasing focus on how to make agent-based technologies safe not only for organizations, but also for consumers and society at large, and this will be good news for the additional players we have already mentioned .
For example, we touched on hyperscalers, but players in the field of AI networks, such as Cisco, also come to mind. They are leading, I would say, thanks to the improvement of the Agentic Ops portfolio, which is specifically designed to solve this problem. You know, how can an organization provide visibility, observability, security—all of these factors integrated into, I would say, a platform that gives them confidence in deploying agents , and therefore creates force multipliers for their existing staff, and also, you know, continue to invest in them and avoid, you know, some of the concerns about isolation; that is, so that organizations do not repeat the mistakes that occurred when building traditional networks, where there is isolation and it is difficult, say, for one agent to interact with another.
And that's where I see the implementation of MCP, as well as new implementations of agent interaction, being of huge importance, and I think that will fuel and drive more agent implementations , and therefore continue to drive the development of AI data centers that we've been talking about. Yes, you talked about AI infrastructure for training and why Nvidia is approaching a milestone that once seemed $6 trillion, while shares are just below record highs after five straight winning sessions.
And according to Raymond James analyst Simon Leopold, Nvidia may still have significant room for growth. His price target is $515, one of the highest on Wall Street. But the main story is not just abou spend huge amounts. New cloud companies are raising capital, and companies like SpaceX are preparing massive investments in AI infrastructure.
And even despite high borrowing costs , . This is a a whole industrial development. . There is also a serious misconcept will be only one winner and everyone else will lose. But the reality can be much more complicated. There can bre many directions simultaneously. And Nvidia is positioned as one of the critical infrastructure providers that connects many of these pieces.
But there is one problem. The bottleneck is no longer simply whether companies wan build led to rising memory prices, and memory is not the only problem. s, cooling, storage systems, and modern components, and it all has focus on GPU demand, the next phas GPUs just as much. And focused on computing power.
Then memory became an important part of this story. The industry is now increasingly focused on transferring information between all these components quickly enough to keep expensive equipment from going down. Think about what this means. You can build huge computing power, you can install a large amount of memory, but if these systems cannot communicate effectively, you a more complex tasks in logic and analytics, architectures become even more complicated.
Different types of computers must continuously interact with each other. This creates another huge opportunity for network technologies , optical connections, switching and related infrastructure. One analyst during the discussion pointed out how significantly the target market for specialized and alternative AI infrastructure has expanded.
The market, once valued at around $14 billion , is now said to be worth around $235 billion. This shows how quickly expectations for AI infrastructure have changed. And it also explains why the possibilities aren't limited to Nvidia. Similar capabilities extend to computing networks , memory and storage, connectivity, security, and infrastructure software.
All of these levels can benefit as AI is integrated into everyday business processes. And there i. The economics of AI are starting to go beyond simply buying more chips. The real question is what economic benefit companies can get from all this computing power.
Comments 0