저자: WilfredFrost외RayWang등—채널/설명란미상 · 외국인
요약 보기
첫 줄 결론: 엔비디아 실적 호조와 AI 수요 확대를 강조하지만, 메모리 비용 상승에 따른 마진 압박은 주요 리스크로 지목됨. [금리·거시] Wilfred Frost는 Jamie Dimon과의 대담을 전하며, 인플레이션이 2%로 떨어져도 10년물 금리는 4~4.5%, 단기금리는 3.25~3.5% 수준이 필요한데 현재 이미 그 근처라 채권 매수 매력이 없다고 봄. 인플레이션은 약 5년간 3% 이상이었고 최근 유가 하락으로 내려왔으나, 1970년대 사례를 들어 재상승 가능성을 경계함. [엔비디아 실적] 분기 매출 약 962억 달러, 데이터센터 매출 약 890억 달러로 예상 상회. 다음 분기 매출 전망 1,080억 달러. 시장 시가총액 하루 약 4,420억 달러 증가(주식 역사상 두 번째 규모). 2028 회계연도 약 17% 성장 시사. [수주·성장 전망] Ray Wang은 엔비디아 수주잔고가 올해 약 7,000억 달러, 2027년 1.2~1.3조, 2028년 약 1.6조 달러로 예상. 고대역폭메모리·네트워킹·포토닉스·서버랙 관련 기업에 기회가 있다고 봄. 가이던스 기준은 1,020~1,100억 달러 수준으로 언급. [제품·소프트웨어] Blackwell 주문이 여전히 강하고, Rubin 등 로드맵을 5~10세대까지 제시한 점을 강점으로 평가. 중기적으로 Physical AI·Sovereign AI 시장(각 1조 달러 규모 추정)과 CUDA·NeMo 등 소프트웨어 생태계 확장을 긍정적으로 봄. [리스크] 메모리 비용 상승으로 엔비디아 매출총이익률 하락 전망. 구글·아마존·OpenAI의 자체 칩 개발 경쟁. 대형 고객 의존도 축소 시도(SpaceX 등 신규 수요) 언급. [출연자 전망] Frost: 시장 전체가 현 수준에서 '최대한 기대할 수 있는' 결과로 평가되는 분위기. Wang: 엔비디아가 AI 컴퓨팅 스택의 큰 부분을 계속 장악하면 AI 사이클 잠재력이 기대보다 클 수 있다고 봄.
다듬은 글
yesterday. And again, this story with the numbers, dig into these numbers, I mean, really dig into them, and I would n't put too much weight on them. They have some stability month over month, but the numbers themselves, I think Kevin Warsh is right when he says: let's see how it all calculates, what they react to and how you weigh and balance them.
And I wouldn't buy, partly because of interest rates, you know, inflation, I mean, even if inflation was 2%, you know, 10-year bonds would have to be at 4.5% or 4% to 4.5%. And the short-term rate should be, you know, at 3.25% or 3.5%. And they are almost at that level today. So I don't see what the advantage is here, even if you think inflation could fall to 2%.
And by the way, it has been above 3% for almost 5 years. And you know, it went up to four, and now, recently, because gasoline prices are falling, it's gone down. It may well fall a little, but you know, as an economic historian, I can't get out of my head what happened after the Great Recession of '74 .
The deficits were smaller, you know, " guns or butter," because the Vietnam War was over, and it was growing, you know, from 3.5% to 5%, 7%, 9%, and 11%. Yes. And you might say, well, oil is a big component. There was an oil crisis in 1973, and then in 1980. Yes , that's true. You know, the unions were more powerful.
All these reasons, but it didn't stop. It was Wilfred Frost, speaking with JP Morgan CEO Jamie Dimon on his “Master Investor” podcast. The full episode, by the way, will be released on Tuesday. The episode lasts about an hour. But Wilfred is here now, at counter 9. Fortunately, he agreed to discuss it.
He is also a Sky News presenter, a CNBC contributor, and our longtime partner. We missed you. We are so happy you are back. Thanks for the invite, guys. Honestly, it always feels like coming home. So, you talked to Jamie Dimon. That's right. And what is your main conclusion? That theme, that phrase he said last week after the earnings announcement , is the maximum one can hope for, these stunning results.
I think he was referring to the stock market in general, as well as the bond market, because we heard in that segment that this is kind of our last chance at this point in time. And regarding the S& P 500 index in general, I tried to ask him a little bit about that . Okay, time to take a detailed look at Nvidia.
All about Nvidia in today's earnings report in our technology roundup. Joining me now is Ray Wang, CEO of Constellation Research, author of "Everybody Wants to Rule the World." Very glad to see you. Ray, when you think about Nvidia, it's not just Nvidia. That's the whole story with artificial intelligence, is n't it?
egins and ends at Nvidia. And this is very important, because for Wall Street, what Jensen says during the earnings conference call and everything related to their forecasts is of great importance. This has a domino effect that increases 10- fold, covering every segment: chips, networks, infrastructure, data centers, software, and virtually the entire AI industry is now in Jensen's hands.
So, where do we start? I mean, from what I see, data centers are going to be good news. Um, the concern might be that they won't be able to build data centers because of politics, or, you know, will there be restrictions on chip shipments to China? There may be some shortage of high-bandwidth memory chips, etc.
We know they are raising prices, and that is actually seen as a positive signal for the tech sector as a whole. They are raising prices because of the costs they face, and that could be good for technology. What, what do you want to start the discussion with? Well, I think the important point when we look at Nvidia as a stock is that we see solid growth.
You will see this in the data center space. They have beaten expectations in 22 out of 24 quarters, five quarters in a row they have beaten forecasts, and while Wall Street is expecting around 92 billion, you are right, that could rise to around 110 billion in the event of a surprise. We are seeing a large volume of unfulfilled orders.
When we talk about these orders, it is not a small volume. We're talking about 700 billion in order backlog this year, 1.2 to 1.3 trillion in 2027, and potentially 1.6 trillion in 2028. This is the domino effect. So anyone who's building high- bandwidth memory, networking, photonics, or even server racks—those are the ones who see huge opportunities right now.
I thought the range from 102 to 110 billion was more of a forecast and a guideline. And so, you know, people will be looking for that . And also, you know , Vera Rubin, and will people still love Blackwell? Will there be demand for this ? So we're also trying to assess what the demand for Rubin and Blackwell will be in the future.
You are absolutely right. This is an excess of 92 billion. The forecast should be somewhere in the range of 105 to 110 billion for people to be excited. Um, Blackwell is still showing good results. You see a lot of orders on the Blackwell side. Um, you'll see the next episode while they work on the implementation.
The beauty and genius of what Jensen introduced was the roadmap. Um, chip manufacturers usually didn't release a roadmap more than two generations ahead. He unveiled a roadmap for the next five, ten generations. And no matter which chip you start with, you can always replace it with a newer one. And that's really the genius of the fact that if you're already with Nvidia, you stay with Nvidia.
Another important point—in the medium term, we are looking at physical AI and sovereign AI. These are also huge markets. Um, and each of them is a trillion-dollar market. And of course, all of this comes together and comes to life in software. When you look at CUDA, NeMo Tron, all the different software assets that they're implementing.
And the deals they're signing, the partnerships they're forming around NeMo Tron and CUDA, are very impressive . Um, not many of them are public, but we are hearing about them from partners and vendors who are already starting this process. There are many public examples that show that they are gaining momentum, even compared to OpenAI or Anthropic, because enterprises need enterprise-grade software , which is exactly what NeMo Tron and CUDA are.
When we talk about NeMo Tron, as well as CUDA as a software for the long term. I mean, what term are we talking about? Is this closer to the present than we think when we talk about the " long term"? I mean, how many months or years are we talking about to see this result? Partnerships are happening today, and I see you'll see it soon.
Nvidia just did something that sounds almost unbelievable. In one trading session, the company added about $442 billion to its market value. This was the second largest single-day increase in market capitalization in stock history. But here's the most interesting thing. The market was n't just reacting to another strong quarter from Nvidia.
He was reacting to what Nvidia said about the future. The company reported quarterly revenue of about $96.2 billion . While data center revenue reached approximately $89 billion. Both figures exceeded expectations, and Nvidia's revenue forecast for the next quarter rose to $108 billion . But the number that really changed the course of the conversation was the long-term forecast.
Nvidia has signaled about 17% revenue growth for fiscal year 2028. This is important because investors have increasingly asked one question. When will AI spending finally slow down? Nvidia responded that at least for now this will not happen. Jensen Huang says that the demand for AI is growing as these systems become economically useful.
More AI use cases mean more tokens, more inferences, and ultimately, more computing needs. And this is where Nvidia still has a huge advantage. But there is another part of the story that deserves attention. Nvidia's growth is starting to extend beyond traditional hyperscalers. So, AI companies, cloud providers, and companies like SpaceX are becoming an increasingly important part of demand.
This is important because it could reduce Nvidia's dependence on a few giant customers . At the same time, Nvidia is expanding its ecosystem strategy . It no longer just sells graphics processing units ( GPUs). It is building a broader AI platform around hardware, networks, software, and, increasingly, the models and developers that use them.
And this becomes especially important as companies like Google, Amazon, and OpenAI develop their own custom chips. Because Nvidia doesn't necessarily have to win every single chip battle if it continues to control a significant portion of the entire AI computing stack. But there is one major risk that investors should not ignore, and that is memory.
Nvidia has warned that rising memory costs are putting pressure on margins, with gross margins expected to decline from current levels. So , the AI boom is creating an unusual situation. Demand is incredibly high, but the infrastructure needed to meet it is becoming more expensive. And so Nvidia's next phase may not be so much about proving that there is demand for AI, but rather about proving that Nvidia can continue to convert that demand into huge profits.
And the market has already seen the headline figures. Now investors are watching for something deeper. How long can this growth machine keep moving forward? After all, if long-term forecasts are even close to reality, the AI cycle may have much greater potential than investors had anticipated . And if you've watched this video up to this point, please like and subscribe to our channel.
Comments 0