저자: JimCramer · 외국인
요약 보기
AI 인프라 투자 확대로 엔비디아의 수요가 폭발하고 있으며, 엔비디아가 데이터센터 AI 가속기 시장을 80% 이상 점유해 생태계 지배력을 가진다고 주장. 오픈AI 등의 수조 달러 규모 컴퓨트 투자 계획과 블랙웰·루빈 등 신규 칩 출시가 매출 성장을 이끈다는 논리. 이에 근거해 2026년 6월까지 주가 500달러 목표가 제시되며, 이는 매출 3배 성장과 고마진 유지 가정에 기반한 단순 계산으로 설명됨.
다듬은 글
Jim Kramer and many others are wildly bullish on Nvidia because AI demand is exploding and Nvidia sits at the center of that storm. Macro players, hyperscalers, cloud builders, governments are committing huge sums to AI infrastructure and that spending is what fuels Nvidia's future revenues. Evidence of that shift is already visible in the numbers.
Open AAI and other model builders have publicly framed compute needs in the trillions over multi-year horizons. That math is a huge driver of demand. Nvidia's share of the market for high-end AI GPUs is enormous. Estimates commonly find it in the 80% plus range, giving it an ecosystem mode. AI is the new industrial revolution.
Think of AI as a five layer cake. Energy plus land plus data centers, the buildings plus chips/infrastructure plus models/lications. Each layer needs investment, sometimes massive investment before the top layer, cool apps can float. The interesting thing for investors is that chips and data center infrastructure are physical and measurable orders, backlogs, capeex plans.
When the top of the stack, Genai models and enterprise applications starts consuming compute like a city on espresso, the lower layers get a growth binge. Nvidia is the company that sells the engines for almost every data center in that cake. If you believe AI apps proliferate, healthcare assistance, drug discovery pipelines, finance algorithms, autonomous vehicles, then the physical infrastructure needed to run them is enormous.
And that's the basic bull thesis. The current scoreboard, Nvidia's numbers, short, factual. Nvidia's fiscal momentum has been enormous. The company reported blockbuster quarterly and annual numbers as AI demand surged with fiscal periods over the last couple of years showing dramatic revenue growth that put Nvidia well into the big cap stratosphere.
That jump in revenue is not vapor. Its bookings, shipments, and customer contracts. Why Jim Kramer and many others are yelling own it. Jim Kramer's rhetoric about Nvidia is loud because he's trying to capture a structural story. The AI buildout is multi- trillion dollar in scope and Nvidia is structurally central to it.
When people like Kramer talk about massive manufacturing and job creation, TSMC partnerships, factories moving compute capacity closer to customers, they're describing the physical footprint of the AI boom, real fabs, powered shells, cooling systems, and supply chains. If you accept the premise that hyperscalers and cloud providers must replace or massively supplement existing compute with new AI tailored hardware because models are getting bigger and inference is getting more demanding, then one of two things tends to happen.
Either multiple vendors suddenly scale and margins compress or a dominant supplier with software and ecosystem advantages like Nvidia reaps outsized profits for years. Kramer bets on the latter. . Open AI and the trillion dollar compute argument. The elephant on stage. This is the meat of the argument.
OpenAI's leadership and some analysts have discussed trillions of dollars in compute commitments over multi-year windows. Numbers like $1.4 trillion have been reported for total commitments related to massive generative AI infrastructure builds. That's the scale of money the industry is talking about to run big models, do inference at scale, and support the new applications popping up every week.
Those are not idle PR lines. Organizations are booking capacity and signing deals. Why this matters for Nvidia? If OpenAI and other model builders keep buying huge compute capacity, someone's chips will run the majority of those workloads. And today, Nvidia's GPUs dominate the high-end AI market. That leads to the everything flows downhill to Nvidia narrative.
Market share and the moat numbers that matter. Estimates across research shops and market writers commonly show Nvidia capturing roughly 80% plus. Some reports higher of the market for the high-end GPUs that power today's large AI models. Remember, it's one thing to own the mid and low-end gaming GPU market.
It's another to own the specialized data center AI accelerator market, and Nvidia dominates the latter. That ecosystem advantage, software, libraries, partner systems, increases switching costs. Imagine every streaming service used a single proprietary codec that only one chipmaker could decode efficiently.
The whole content ecosystem would depend on that chipmaker's hardware. That's sort of the Nvidia situation for the high-end AI stack. Blackwell, Reuben, H series, the hardware story. Nvidia hasn't just sold GPUs, it's iterated fast. The company's newest architectures, often marketed with dramatic names, have been designed specifically for AI workloads, inference, and training.
And the launch cadence coupled with software stacks keeps enterprise customers coming back instead of building an internal alternative overnight. When new chips, Blackwell family, Vera, Rubin, etc. arrive and deliver big performance gains, customers pre-order and lock in capacity for future models. That's how fiscal forecasts can jump quickly.
If chips double performance while staying power efficient, customers buy more because performance equals lower total cost of ownership for a given workload. It's a virtuous loop for Nvidia. Shipments and pricing power rise margins stay healthy. If that loop holds, consensus revenue numbers can be beaten quarter after quarter.
The hyperscaler circularity and OpenAI's funding drama. You'll recall a media narrative last year about circular deals. Open AI taking investments from companies which then buy cloud and compute from the same ecosystem that created hand ringing. Who pays for all this? The short answer, multiple sources, private investments, hyperscaler budgets, enterprise customers will fund infrastructure.
Open AAI has also raised very large sums including major investments from big names and strategic partners that help make their commitments credible. Even so, questions about who foots the longerterm bill linger, and they create episodic volatility in the AI supply chain names. The key point for Nvidia, whether the money is circular or not, someone is spending on chips and data centers.
And today, many of those spending decisions point to Nvidia hardware. The math behind $500, a simple back of envelope model. Okay, now to the juicy ask. How do people arrive at targets like $500 by mid of 2026? Strip away the hype and its math. Revenue ramp assumption. Analysts in some models assume Nvidia's data center revenues and total revenues grow dramatically.
Sometimes modeled to reach hundreds of billions on an annual basis within a couple of years. Numbers like $200 billion or more for FY26 show up in some analyst sets. If revenue triples and margins hold, profits explode. Margin assumption. Nvidia's data center GPUs historically have much higher gross margins than commodity chips.
Assume a blended high margin 40 to 70% depending on mix EPS and multiple multiply projected net income by an earnings multiple. Markets have been willing to pay premium multiples for high growth tech with durable modes. That gives a market cap divide by shares outstanding and voila lofty per share target in .
Comments 0