저자: JimCramer · 외국인
요약 보기
영상은 CrowdStrike CEO 인터뷰 중심으로, AI 기반 보안 수요와 파트너 확장(Accenture, EY 등)을 강조하며 CrowdStrike가 취약점 탐지만이 아닌 전체 보안 플랫폼으로 경쟁 우위가 있다고 주장. 보안 시장은 플랫폼 기업 중심 과점 구도로 재편될 것이라는 전망. 이후 Nvidia 관련 분석으로 넘어가, AI 수요는 공급(전력·데이터센터 인프라)보다 빨라 병목이 칩에서 전력으로 이동 중이며 인프라 확장 수혜가 지속될 것이라 주장. 구체적 목표가(2200달러)는 제목 외 근거 자막에서 확인되지 않음.
다듬은 글
It is. Well, when you when you think about Glassdoor when it came out, you have only a handful of companies that have access to Me those. So, you have a lot of customers their boards, their CEO saying, "What what can we do?" I mean, our CEO is calling me in the middle of the night, Fortune 10 companies going Me those, you know, help us.
What can we do? How does all this work? And and you know, how do we make sure we're secure? So, we're in the middle of these conversations and the opportunity is so large that you need to create a coalition around this, the Accenture's, the EY's of the world, etc. where they can work using our technology to help customers understand where there is where their exposures are.
We're finding millions of exposures in companies that we go into. Those need to be prioritized and you need to understand what mitigating controls like Falcon are in place to be able to prevent against those. So, the great opportunity here is we have our partners drop into either customers or non-customers that we turn into customers because they're using our technology as part of this AI readiness assessment.
So, it's really a multiplier effect for us. Don't aren't you surprised that when everything first broke, even in the winter when we realized all the different vulnerabilities, that somehow all this stuff was There There was a belief that it was bad for CrowdStrike when I know I know when I first saw it, I said, "Oh my god, Bonanza." And yet all I read for so long was this is the end of CrowdStrike.
What do you think went wrong with that narrative? How did people just get it completely backward? Well, Jimmy, you have to look at security. There's many different facets and finding vulnerabilities in source code is one facet of it, okay? And that's what the LLM models do pretty well. But then you have to look at the overall purpose of security.
It's to stop the breach. You have to be in line. You have to be You have to get it right, you know, the first time you see a threat. You've got to be able to create data that allows you to understand what's happening and train that data. And we have a massive proprietary data mode from, you know, over a decade of doing this.
So, at the end of the day, we can use these LLMs to be more successful and and and have greater impact with our customers, but it's only one aspect of what you need to do to build a security platform like Falcon. So, when this came out, I think a lot of people were confused that if you got rid of all the vulnerabilities, that the world would be safe.
That's not the case. You can get rid of all the vulnerabilities in the world and you would still have break-ins because you have things like identity theft. And what I mean by that is your credentials, your username, your password, your tokens get stolen, people log into a system and steal all that data.
So, there are many ways to get into a system. Exploiting a vulnerability is just one of those, and you're never going to get rid of all of them, and you're still going to need a full life cycle system, a system of record like Falcon to be able to stop those breaches. And I look, I know that I'm I've been limiting my questions to this notion of what happens with with the anthropic emitters.
There's been no diminution of the just plain vanilla, I'm going to wreck your company, pay me kind of things that that CrowdStrike does for a living. Absolutely not. You know, Jim, we have talked extensively about extortionware and ransomware, right, where these the cyber gangs will basically get into a a large company or even small companies that don't have the level of protection.
It's really shameful. They get in and they extort these folks, and you need security up and down the the entire business stack, right? Whether it's a small business or whether it's a large enterprise. And we're only seeing more and more of these sort of attacks. What AI has done is it's created more adversaries with greater sophistication because they're leveraging the models themselves.
So, what does that mean? It means more tailwinds for companies like CrowdStrike. Okay, last question. You There were so many people in cybersecurity when I first started talking to you. I kind of feel like that it's gotten so big and the threats are so horrible that you need a complete suite. And the only one other than you that has a pretty complete suite would be Nikesh Nikesh Arora's company, Palo Alto Networks.
Is it just you and them? Is it just a duopoly now? Well, I think the the platform vendors are ultimately going to win. I would Yeah. So, that's so be semiconductors though is not a function of war or geopolitical issues. That's a function of AI. I would say that AI is even the dividing line for geopolitics.
And so, not just for the markets, but also if you want to be on the right side of the economy and on geopolitics, you better be also the leader in AI. So, I think it transcends a lot of different dimensions. Does the euphoria in the market around all these IPOs make sense to you? So, I think right now for me actually a more interesting question than maybe the flow question that you're alluding to is really how disruptive AI is to incumbents.
When we look back at at paradigm shifts in the past, we have seen leadership in the market changing. You think about the mainframe era for example. IBM was the defining company of the mainframe era. Then, it passed the baton to Intel and Microsoft. And then of course, Apple took the lead with mobile computing, the first $1 trillion company.
Right. And I think the question now is how disruptive AI is to those market leaders. Okay, and what what do you what what do you think about that though? Yeah. So, we have seen what happened in software, but I think AI coding is so disruptive to software and I think agentic AI could very similarly be as disruptive to the internet.
Because if more and more agents are executing on our behalf, the power balance can shift away from internet to those agentic platforms. Have you changed internally at the firm itself? How you're operating, how many people you're hiring, who's on your team, everything else as a result of AI yet? We have.
Actually, one of the thing that I think that is so under appreciated by the market is how little the traditional factor buckets of country, sector, style, and size capture in terms of investment returns in the equity markets. So, we actually formed a team that is focused on AI 2 years ago. It's a global team.
It's not a sector team to understand the investment opportunity and then to come up with portfolio &'s biggest problem isn't the competition. It's keeping up with demand. For the years, investors worried about one thing when it came to Nvidia, that is competition. Would other chips company will catch up?
Would hyperscalers replace Nvidia with custom silicon? And would a new startup disrupt the market? Today, Nvidia faces a completely different challenge, keeping up with demand. And according to Jensen Huang, the industry's biggest bottleneck may be not even be chips anymore. It's power. Think about that for a moment.
Not customers, not applications, not adoption, that is power. The AI industry has advanced so quickly that electricity infrastructure is struggling to keep the pace. Data centers require enormous amount of energy. And every new AI models require more compute, every new AI agents require more inferences, and every new AI power services require additional capacity.
And all of it ultimately depend on power. That's why conversations throughout the technology industry are increasingly shifting toward the infrastructure. Power generation, grid expansion, data centers construction, cooling system, networking, memory. These are becoming just as important as the chips themselves.
In many ways, the AI race is transforming into the an infrastructure race. That's incredibly bullish for companies positioned at the center of ecosystem. Nvidia remain one of those companies. The reason is simple, no matter which AI trend emerge next, more compute is usually the answer. Is it the AI more compute?
AI assistance more compute? Physical AI more compute? Every road seem to be lead back to the infrastructure. That's why many investors believe the market is still underestimating the duration of this cycle. A years ago, many people expected that AI is spending too slow. Instead, the spending is accelerated.
Today, many of rather than reducing them. They aren't acting like a company preparing for the slowdown. They are acting like a company trying to secure a square resources before competitor do. That's a very different distinction and it's the one of the reason Nvidia ecosystem continue expanding. The company is no longer just a GPU supplier.
It's becoming an infrastructure a software platform, a networking provider, a system architects and AI factory builder. The deeper AI penetrate the economy, the more valuable that ecosystem became. What fascinating is that many investors remain focused on quarterly stock momentum while ignoring the larger transformation taking place under the arm.
Artificial intelligence is steadily becoming part of everyday life, not just inside research labs, not just inside the cloud providers, everywhere. Consumer devices are becoming AI devices. Enterprises software is becoming an AI software. Manufacturing is becoming an AI driven. Healthcare is becoming AI assistant and eventually many physical product will contain intelligent agents.
All of those trend are create additional demand and unlike the previous technologies angle, demand is arriving before enough infrastructure exist. That's why they they are still shortages and that's why data centers capacity remain constrained.
댓글 0