저자: JayGoldberg · 외국인
요약 보기
엔비디아는 AI 스토리가 이미 전부 가격에 반영됐고 Blackwell 물량이 매진돼 상방 여지가 작아 매도 의견을 제시. 관세보다 미중 AI 경쟁과 수출규제가 핵심 변수이며, 빅테크 capex 확대는 좋지만 이미 반영됐고 CFO 중심의 AI 투자 수익성 점검이 늘어 조정 가능성을 지적. 반면 브로드컴은 AI 스토리가 가격에 충분히 반영되지 않았다며 매수 의견. 딥시크 충격으로 브로드컴 상승분이 반납된 점도 언급.
다듬은 글
Seaport Research hitting Nvidia with the rare sell rating as it initiates coverage on semiconductors. Seaport noting AI is bringing about a secular shift in the industry. Jay Goldberg, senior analyst at Seaport Research Partners joining us now to discuss this. Jay, it is great to have you here. So, you initiate Nvidia with a sell.
Uh you might, Jay, per Bloomberg be the only sell on on the street. Explain the rating rating to us, Jay. Great. Great to uh thanks for having me. So, put simply, I think the AI story, we've been talking about it for years, we all know about it. It's fully priced into Nvidia right now. They're a good company, they make great products, but they've said already their newest line of products, their Blackwell line, is sold out for the year.
They're capacity constrained. Once you're sold out, the the room for upside is is pretty minimal. The bias shifts towards the downside, right? There's not much they can do to get above where consensus is, but there are a lot of things that can trip them up. There's lots of small things, snafus, there's a lot of friction in the system, and I think that uh we should be concerned about those, and that, you know, I semiconductors are cyclical, and it's important to remember that.
Um Jay, um what about this idea that there might be some sort of potential tariff or export control relief for Nvidia um that wouldn't necessarily affect Blackwell, but could affect some of its other chips? Well, I I think that there's two sides of this tariff story for Nvidia. One is their cost of importing things from China.
A lot of the supply chain comes from Asia. Most of the analysis we've seen shows that that impact is fairly minimal. So, that the the hyperscalers, the Googles and the Amazons of the world, can continue to buy Nvidia chips and import them into the country at at essentially, you know, no no tariff or very low tariff.
The flip side of that is what's going on in China. China would very much like to get their hands on more Nvidia chips, um but that seems very increasingly difficult in this geopolitical environment. So, it's less about tariffs and more about this perceived rivalry between the US and China's AI race.
Uh and I I I don't see that lightening up anytime soon. And even if that does happen, uh you know, it's it's still fairly minimal. I think we'll still there will still be some restrictions, so it won't be a an a quick snapback. Jay, you know, on the big tech companies that have reported, it's interesting on capex, Jay.
I mean, and you've watched, obviously. I mean, you saw Alphabet reiterate, Microsoft reiterate, uh Meta actually raised. Uh that would seem to be all good news for Nvidia, Jay. So, I guess you're saying, yes, it is good news, but that that simply is priced in at these levels. That's right. It's priced in for these levels, and I think it's important to remember what they're talking about is capex for this year.
It's a little early to be talking about capex budgets for next year, but in my conversations I'm having with everybody is there's a growing scrutiny being applied to AI spend. We've had some really, really big numbers this year from all the hyperscalers. I think it's we're starting to see signs that, you know, I don't think anyone's going to cut drastically, but there are certainly signs that the finance team is starting to ask questions.
A year ago, it was all AI all the time, but now the the CFO has entered the chat and is starting to ask like, what what's the return on this investment? And just to be clear, I'm a believer in AI. I think there's a lot of potential there. I just think it's going to take us a few years as an industry to really find those compelling use cases.
And as we do that, there's going to be, you know, a a sort of gravity asserting itself, and people are going to ask, what's the return on this investment? And I think that is that that sets us up for a potential for a pause, or at least a slowing. Jay, even as you have the sell on Nvidia, you do have a buy on Broadcom.
And of course, Broadcom has gotten a lot of attention as being um a company that could benefit from more companies sort of doing custom chips or contract-made chips. Um but when you talk about stuff being priced in, I mean, Broadcom is up about 60% over the last year. Nvidia's up about 36% or so. So, how much of that outlook is priced in?
So, I I think this I what I told my sales force is that the AI story is fully priced into Nvidia, and it's not priced into Broadcom. Broadcom has a lot of other pieces going on there beyond the AI story, their networking business, their software business, but I don't think the street entirely believes their AI story.
And we actually saw this. They reported earnings in December, and for the first time they laid out their opportunity for AI, and it was large. It was a massive opportunity. Anytime Nvidia loses share today, it's to one of the hyperscalers doing their own chip, and Broadcom makes a lot of money helping them bring those chips to market, right?
So, when they laid out their AI story, Broadcom was up 20, 25% on that, right? And then in January, we saw the deep seek story come out, and everyone had a momentary panic that we weren't going to need GPUs anymore. Uh and Broadcom essentially gave up all those gains, right? And I think I think there's still a a lot of show me built into the Broadcom story around AI.
And you know, the street doesn't believe it. I do. Jay, thanks a lot. Appreciate it. Thank you.







댓글 0