저자: TomLee · 외국인
요약 보기
AI 인프라 슈퍼사이클이 금리 상승에도 지속되며 기술주가 시장을 끌어올릴 것이라는 주장. 칩 수급비율 13:1, 하이퍼스케일러 투자 확대, 약 5,000억 달러 규모 사모대출 자금이 거론됨. 엔비디아(저PER·현금흐름 근거), 스페이스X(10GW 컴퓨팅·Starlink 마진 근거), Vertiv(수주잔고 150억 달러·수요 근거)를 최선호 종목으로 제시하고 조정 시 매수 기회라고 강조.
다듬은 글
I don't think AI will lead to significant job losses. The supply-demand ratio for chips is now 13 to 1 , and you've seen it with Nvidia and AMD. And I believe that investors continue to underestimate the scale and scope of this AI revolution. But investments in AI create inflationary pressure on memory and energy consumption.
Um, these are bottlenecks. The reality is that raising Fed rates will not slow down spending on AI. It is technology that will push this market up, and I think that is what is happening in the spotlight right now. What you just heard from Tom Lee and Dan Ives is not just typical Wall Street optimism.
This is the reality of the historic $4 trillion AI infrastructure supercycle unfolding before our eyes. Despite aggressive Federal Reserve policy, ongoing rate volatility, and macroeconomic uncertainty, corporate spending on AI and hyperscaler profits have completely exceeded expectations. Double-digit growth in cloud computing by large companies is directly funding an unprecedented scramble for computing power, while private lending giants are creating a pool of capital of about $500 billion to support megaprojects.
This is most clearly seen in the example of SpaceX's massive $40 billion Apollo-led financial package. A stunning deal designed to funnel tens of billions straight into high-performance GPUs and the creation of next-generation supercomputers like the Colossus 2. Macroeconomic headwinds simply cannot stop this structural wave.
But to truly benefit from this debt-fueled expansion, you need to look beyond standard hyperscalers and monitor the physical supply chain. Today we will look at three prominent market leaders who are at the center of this structural boom. First, a dominant engine that powers global AI computations and orchestrates its own demand through strategic partnerships with stocks.
Second, a major space and computing giant using private credit to build unrivaled scale supercomputers. And third, the critical infrastructure support that provides high- density liquid cooling and power solutions, without which these multi-billion dollar data centers cannot operate. Given their structural monopolies, accelerating earnings growth, and multi-year orders, their current multiples make them extremely cheap for long-term investors, turning any market pullback into a rare and excellent buying opportunity.
Let's break them down. First on the list is Nvidia, ticker NVDA, the dominant engine driving the AI revolution, which is strategically transforming itself from a hardware provider to the main financial foundation of the entire AI ecosystem . To overcome capital constraints that could slow data center expansion, Nvidia is using its massive free cash flow, which reached a staggering $127 billion over the past 12 months, as strategic support for large-scale infrastructure projects.
By providing financing platforms alongside major private equity firms and offering up to $250 billion in guarantees for next-generation data centers , the company is scaling global AI hardware spending to $500 billion while maintaining a dominant gross margin profile of 72-73%. This means that every dollar allocated to credit guarantees can generate equipment sales and gross profit that far outweigh the potential credit risks.
From a product and demand perspective, growth signals remain extremely strong . While short-term supply constraints are shifting across the semiconductor manufacturing chain, major partners like Micron have already sold out their high-speed memory capacity through 2027 at higher prices, specifically tailoring custom solutions for Nvidia's next-generation architectures .
In terms of cloud deployments, major partners like Oracle shipped over 300,000 GPUs in a single quarter, reflecting a huge backlog that directly feeds Nvidia's pipeline. Beyond hardware, strategic moves like the $ 12.9 billion acquisition of Hugging Face and the launch of enterprise AI security platforms are creating a hugely compelling software ecosystem that is tying millions of developers to Nvidia's architecture.
On the financial front, consensus revenue forecasts are rising to nearly $692.21 billion by fiscal 2028, supported by management's unprecedented increase in share repurchase authorization by $150 billion . With earnings growing faster than the share price, valuation multiples have fallen to historic lows , around 15 to expected earnings , creating a significant gap between market caution and the fundamental acceleration that positions Nvidia perfectly for long-term growth.
Moving from silicon-level to physical scaling of orbital computing, our second stock is SpaceX, ticker SPCX, a company that is fundamentally reinventing itself as an integrated platform for orbital communications and massive AI computing. While SpaceX is historically known for its launch dominance, its computing power trajectory is accelerating at an impressive pace, increasing from 1.4 gigawatts of installed computing power at the end of the second quarter to a projected 10 gigawatts by 2027.
This explosive capacity buildout is already delivering incredible revenue momentum: AI revenue grew to $2.6 billion in the second quarter alone, alongside $1.6 billion in additional hosting revenue. Importantly, the economics of these computing deployments are proving to be extremely rapid , demonstrating payback times of less than one year and laying the foundation for a business segment that, at scale, could conservatively generate $64 billion to $80 billion in annual high-margin revenue.
This aggressive expansion is fueled by SpaceX's major broadband moneymaker—Starlink, which generated over $11.39 billion in revenue and $7.17 billion in adjusted EBITDA with an operating margin of 63%. Serving approximately 12 million subscribers and rapidly acquiring strategic assets, including the $8 billion acquisition of a nationwide portfolio of 800 MHz spectrum , Starlink provides a highly profitable, self-sustaining foundation that supports SpaceX's massive capital investment .
Additionally, with Starship's full commercial reusability , launch economics are rapidly improving , opening up long-term potential for orbital data centers using constant solar power and low-latency optical laser networks . While the company's capital intensity remains enormous—over $28.48 —its unprecedented vertical integration across space transportation, satellite communications, and computing infrastructure creates a structural barrier that is nearly impossible for competitors to replicate .
As operating leverage kicks in and computing power grows to 10 GW, SpaceX offers investors incredible long-term prospects for multi-trillion dollar growth, aimed at taking advantage of the next phase of AI and communications infrastructure. Completing this powerful infrastructure triad is our third company—Vertiv Holdings, ticker VRT, the undisputed foundation of the physical infrastructure that ensures the operation of huge AI supercomputers without overheating or grid failures.
While headlines about AI are dominated by silicon chips and launch capabilities, each installed accelerator creates unprecedented heat density and energy demands that traditional facilities simply cannot handle. Vertiv operates at the very epicenter of this physical bottleneck, specializing in combined high- density liquid cooling, power distribution, and precision thermal management.
With about 85% of revenue coming directly from data centers, Vertiv occupies an elite position as an end-to-end electrification and cooling partner, even collaborating with Nvidia on an 800V DC power architecture for new computing platforms. The structural demand driving Vertiv's growth is supported by large-scale long-term capital investments.
With multi-trillion dollar data center spending projected by the end of the decade, Vertiv has built an impressive $15 billion order backlog, providing high revenue visibility for many years to come. This surge in complex installations has opened up significant operating leverage, increasing gross margin to 37% and ensuring rapid generation of operating cash flow.
Additionally, Vertiv's growth is increasingly fueled by its high-margin service segment, which grew by over 28% as customers rely on continuous temperature monitoring and closed-loop liquid cooling systems that minimize water waste. In terms of competition, Vertiv is part of a narrow global oligopoly with Schneider and Eaton, but stands out as the purest player due to its complete focus on data center environments.
Recent strategic acquisitions, such as the $1.45 billion purchase of Utility Innovation Group to expand into the microgrid sector, further strengthen Vertiv's role in addressing the critical electricity affordability crisis. While challenging project execution can sometimes cause minor quarterly shifts, raising the annual revenue forecast to $14 billion and accelerating operating profit by nearly 60% demonstrates phenomenal internal momentum.
As an infrastructure leader with growing margins and a huge multi-year order backlog, Vertiv offers an attractive value proposition whenever short-term pullbacks squeeze its valuation multiple. As private debt, orbital computing, and physical energy converge into a multi-trillion dollar supercycle, these three market leaders offer the most direct and reliable way to invest in AI development.
Which of these options will you add to your watchlist ? Let us know in the comments below. Like, subscribe, and see you in the next detailed review.
Comments 0