저자: 짐크레이머 · 외국인
요약 보기
짐 크레이머는 이 영상에서 AI 확산으로 인터넷 트래픽이 급증하면서 콘텐츠 전송(CDN) 기업들이 수혜를 볼 것이라는 주제를 중심으로 이야기한다. 그는 클라우드플레어(Cloudflare)는 올해 이익의 279배에 거래돼 비싸고, 파스틀리(Fastly)는 PER 50배에 변동성이 커서 신뢰하지 않는다며 지금은 매수를 권하지 않는다고 밝힌다. 반면 아카마이(Akamai)는 예상 매출의 16배 미만으로 싸고 앤트로픽(Anthropic)과의 파트너십 가치가 크다며 지금 강하게 추천한다. 아카마이는 올해 비용 증가로 부진하지만 곧 개선될 것이라고 본다. 주요 콜러 질문 부분에서는 액손(AXON)에 대해 경쟁 심화로 매출총이익률 압박이 예상되고 PER 54배는 부담스럽다며 당분간 휴식(관망)을 권한다. 인트라도(Intrado)는 솔직한 기업이라 호감을 표시했고, 코발론(Covalon)은 재무구조가 마음에 들지 않아 소액만 보유하라고 했다. 인터넷 트래픽 증가 흐름 속에서는 콘텐츠 전송 기업들에 관심을 가질 만하다고 덧붙인다. 이어 마이크론 실적과 관련해 반도체 및 AI 관련주가 주 초 급락한 상황을 언급한다. 루멘텀(Lumentum) CEO 인터뷰에서는 미국 행정부가 중국산 광트랜시버 수입 제한을 검토 중이라는 소식이 다뤄졌다. 중국이 광트랜시버 공급망의 약 70%를 차지하며, 규제가 잘못 시행되면 AI 데이터센터 확장 속도가 둔화될 수 있다는 우려가 제기된다. 루멘텀은 AI 데이터센터 인프라에 의존하지만 광 제품 전환 흐름으로 충격이 제한적일 것이라고 주장한다. 마지막으로 스페이스X의 실적과 엔비디아 관련 논의를 소개한다. AI 컴퓨팅 확대는 GPU뿐 아니라 전력, 냉각, 네트워크, 메모리가 함께 갖춰져야 하며, 전력 및 냉각 병목이 주요 제약이라는 점이 강조된다. 엔비디아는 이 수요 흐름의 중심에 있다고 평가된다. 크레이머는 전체적으로 AI 수요는 견조하지만 종목별 밸류에이션과 경쟁·규제 리스크를 구분해 접근해야 한다는 입장을 보인다.
다듬은 글
I love Cloudflare, but it's selling for 279 times this year's earnings . On the other hand, this company's shares have always been expensive, and if that scared you away, you would miss out on huge profits. However, if you want to buy Cloudflare now, I recommend waiting for a significant market pullback.
Given Iran and the Federal Reserve, I'm sure someone will give you that opportunity. Fastly has a more reasonable valuation, trading at a price-to- earnings multiple of 50 times, but that's still a lot, and the stock is prone to wild volatility. I'll be honest, this is the company I trust the least.
So again , I urge you to wait for a more favorable entry point. But I think Akamai is worth buying right now because these shares are cheap. They sell for less than 16 times projected annual revenue, and you essentially get a partnership with Anthropic, which I think is huge, for free. Sure, they're having a bad year because of all those expenses I mentioned, but things will change for the better soon.
Here is the main conclusion. One of the side effects of the AI revolution is that it is already creating a lot of internet traffic, and someone needs to help us keep the network running smoothly. This is where Cloudflare, Fastly, and Akamai come to the rescue. Although at this point, Akamai is the only company that is cheap enough that I can strongly recommend it.
I want to take a few calls and I'll start with Linda from Pennsylvania. Lindo. Congratulations, Jim. May God protect you and Lisa. You are like a family member to me , and when you are on vacation, I love all your extra reviews. I hope Norman and the dog are doing well. Jim, I'm calling about an IPO that I've been trading since inception; I made money, but since April the stock has fallen by 26%.
I know about the debts, and yesterday you had guests from CC, Michael and Trader. So I'm wondering if I should take a profit and wait for something better, or wait? Well, Linda, first of all, thank you for your kind words and for what you said about Lisa. You know, she's the one who keeps me from overworking.
She says, "You need to take a vacation, Jim. You really need to do this." So that's great advice. I really like Michael and Trader. The stock is still almost double its initial price. Yes, that's twice as much as when the company went public. I have tremendous faith in him. Stocks really lose a lot of money.
So the question is , do you have the stamina for it? If I were you , I would say, " You know what? I want a small stake in Covalon, a small one, because I don't like their balance sheet ." But that's part of the business. But I really like Intrado. He is a straightforward person. Robin from Rob. Boo-ya, Jim.
Boo-ya, Rob. Jim, what are your thoughts on AXON? Thank you. Okay, I like Rick Smith. Anyone who has been watching the show for a couple of decades knows that I think Rick Smith is great. I really think there is real competition in Axon. They have a good product, but I have to say that when there is competition, it means pressure on gross margins.
When there is pressure on gross margins, the stock cannot be sold at a 54-to- earnings multiple. So the answer is: I'm taking a break from Axon right now. I would have been excited about Covalon, but I was a little too negative. I think Covalon is the one Jensen wants. He believes this is the best developer.
I'm not going to attribute this to Jensen, at least Intrado said that Jensen said it was the best option. With AI generating all this web traffic, I think these content delivery players are worth paying attention to. I like Akamai. Let's remember when, a year after the IPO, my company firmly proved that it deserved to be public.
I'm looking at these stocks. Then Micron reported earnings last night, and today the stock, well, at first it didn't show much movement. I explain why I supported this name during a period when today was really bad. And we call it "rapid fire" in today's Lightning Round. So stay with us. Based on what we see, we certainly provide a lot of the equipment that goes into these data centers.
This is equipment that they have to spend money on long before deploying data centers. And from what we see, their costs are increasing. The Trump administration could put a damper on, as I understand it , the entire industry as a whole . They are considering whether to impose restrictions on imports of Chinese optical transceivers.
And for those who don't know, they are critically important for the operation of hyperscaler data centers because they transfer data very quickly. China dominates a large part of the supply chain, and the administration, as you know, Mike, wants to bring manufacturing back to the United States. You're already building a facility in North Carolina, but let me ask: Will the ban on Chinese imports help Lumentum, or is this a case of being careful what you wish for?
Very good question, Liz. Hmm, that's a double- edged sword. We would certainly benefit from this . So these transceivers that you just mentioned are the backbone of connectivity inside the data center. You have all these computing blocks. How to connect them together? Usually lasers are needed. We need devices that will transmit data using light.
And these transceivers are essentially a way to speed up communication between all the different subsystems in the data center. Today, 70% of these transceivers are supplied to China. We manufacture transceivers, but we are a small player. There are one or two other American companies that make them.
Together we occupy 30%. So if these rules are handled incorrectly, on the one hand, we would benefit, but on the other hand—I think it would hit the AI sector, and you could see a slowdown in the pace of development. So we want to work with the government and make sure that these rules are implemented in the most appropriate way.
You know, this week the AI sector, as we've said, has really taken a hurricane-like hit. I mean, your own stock fell about 21% from Monday to Thursday. Wall Street, however, loves revival stories, but JP Morgan said this week that it shouldn't be expected anytime soon, especially after semiconductor stocks plunged.
The bank warns that investors who are used to buying on dips are disappearing. Michael, you 're selling your products as fast as you can produce them, but the stock, yes, has fallen sharply this week. However, over the past year, you have shown exponential growth. What are investors missing or ignoring that you don't see in your order books?
What can you say now about the future of Lumentum that might change their minds? Yes, I think, Liz, you are following this story very well. Our numbers today are based primarily on the development of data center infrastructure for AI. So everything you talked about in the AI sector this week would impact us.
There is no doubt about it. But the difference between what we're doing and what someone else is doing, you just mentioned Gok Tan, is that we're not as dependent on building AI data centers in the future. In data centers, there is a shift from electrical products, such as those manufactured by Broadcom, to more optical ones.
So, we may see a world where data center construction slows down significantly, but for Lumentum, it will have minimal impact because Before we begin, on its current pace. But the new numbers from SpaceX raise another question. What if the AI industry doesn't lack demand? What if it lacks the energ computing for AI, capacity.
Meanwhile, reports of GPUs waiting for power have sparked new debate s reports from the biggest tech companies matters so much. SpaceX figures that attracted the attention of investors. 's higher than the roughly $12.7 billion analysts had expected. Its space biggest area of interest has been the busines the ignificant increase from the prior year.
These numbers are important because this , supplying AI puts Nvidia right at the center of events. I strik SpaceX plans to reach a much more ambitious goal by taken as a management forecast, not increase its solid computing capacity with the goal of reaching about 10 gigawatts of additional capacity by 2027.
Musk also discussed the grand ambition of 20 gigawatts of additional power and cooling, acknowledging the complexity of the task. Why is this important to Nvidia? Because any major expansion of AI computing requires more than just software. You need processors, memory, networking, pulls off will have a significance that goe points during SpaceX's financial report was the discussion of energy and cooling.
Musk suggested that the biggest problem isn' operate . You can't just install 1000 GPUs and start making money immediately. The building requires sufficient electricity, and cooling systems must also be installed. Network equipment needs to be installed, and memory and servers need to work together.
The entire system will need to be tested and connected to the network. If even one important component is missing, part of the computing power may be idle. indicate some of its facilities had enough power to operate them.
Comments 0