저자: DanIves · 외국인
요약 보기
Dan Ives는 AI 열풍이 '밤 10시 30분 파티'처럼 초기 단계라며 향후 2~3년간 기술주 강세장이 이어질 것이라 주장한다. 근거로 TSMC 매출 전년비 40% 증가(330억 달러), 고성능컴퓨팅 매출 비중 절반 이상, AI 칩 수요 대비 공급 약 10:1 격차를 든다. 이어 TSMC, ASML, 마벨(MRVL)을 AI 인프라 핵심 수혜주로 소개하며 버블론을 일축한다. 영상 후반은 광고성 멘트(무료 금 투자 가이드 링크)가 포함되어 있다.
다듬은 글
10:30 p.m. in this AI party. It started 900 p.m. and that party goes to 4:00 a.m. And the reality is like, look, this is going to be a 2 to threeyear left in this bull cycle for tech. Look, haters are going to hate. Bears will say it's a, you know, this is a bubble. I could just tell you this is a fourth industrial revolution.
We are just in the early days of this playing out. I think those numbers and we're going to see throughout tech earnings, it's going to be validation for the AI revolution. And I think the bears, look, when the bears are in hibernation mode, they can't see AI in the spreadsheets. Famed analyst Dan Ies just dropped one of the most audacious predictions we've heard all year, and it's setting the stage for the next exciting chapter of this bull market.
In his most recent interview, he famously declared, "It's only 10:30 p.m. at a party that goes until 4:00 a.m. What's the takeaway? The AI frenzy isn't winding down. It's barely gaining momentum." Ives argues we're just at the beginning of what he terms the fourth industrial revolution. A massive multi-trillion dollar growth period set to overhaul not just technology but the entire global financial system.
This is completely different from the dot era's burst bubble. Back then investors were chasing hype and untested promises. Today we are witnessing the actual tangible and digital foundations of artificial intelligence being constructed in real time. Data centers, microchips, and cloud networks that will serve as the essential framework for our intelligent economy.
As IV sees it, the doubters are still buried in old data while the true believers are already capitalizing on tomorrow's wealth. And the proof is everywhere you look. TSMC's latest earnings reports show that demand for AI chips is outstripping available supply by an astonishing ratio of nearly 10:1.
Powerhouses like Microsoft, Amazon, and Google are pouring unprecedented amounts of money into AI infrastructure. Meanwhile, utility providers, cloud platforms, and chip manufacturers are struggling to keep pace. With inflation cooling off, interest rates becoming more stable, and cash flow returning to high growth sectors, the conditions are perfect for a major 2 to three-year tech bull run.
In this feature, we're going to zoom in on three companies positioned directly at the core of Dan Ies's forecast. These aren't just firms dabbling in the AI revolution. They are the ones literally building it from the ground up. They are laying the digital railways for the coming decade and stand to be the single biggest winners of this unstoppable AI growth cycle.
We kick things off with the company Dan Ies dubs the heartbeat of the AI revolution. Taiwan Semiconductor Manufacturing Company Limited or TSM. If the brilliant minds at Nvidia design the brain of artificial intelligence, TSMC is the one who manufactures it. The company fabricates over 70% of the world's most sophisticated microchips, giving it undeniable command over the semiconductor supply chain that powers every single AI system globally.
Fun fact, TSMC pioneered the dedicated foundry business model, meaning it focuses solely on manufacturing chips for other designers, which is why it's the partner for nearly every major tech brand. TSMC's recent quarterly results were nothing short of breathtaking. Revenue rocketed 40% year-over-year to $33 billion, easily surpassing even the most optimistic analyst predictions.
What fueled this enormous growth? High Performance Computing, HPC. This division, which handles the processing power for AI models, large language systems, and massive scale data centers, now accounts for more than half of TSMC's total sales. Demand for cuttingedge chips isn't just high. It's absolutely exploding with production unable to keep up with the 10-1 demand gap.
The company's next breakthrough, its two nanometer chip technology, is already in the final stages of development, promising unprecedented power efficiency and speed. Production is slated to begin later this year with a full-scale roll out by 2026. Every major tech innovator from Apple and Nvidia to Amazon and OpenAI's hardware partners is already lining up to secure a spot.
Furthermore, TSMC is strategically reducing risk by expanding its manufacturing footprint worldwide. It's investing over $ 160 billion in new factories, fabs, across the US, Japan, and Europe. This monumental shift lessens its geopolitical risk while simultaneously reinforcing partnerships with Western governments.
In the AI economy, TSMC isn't just one chip company among many. It is the very bedrock upon which every other AI player depends. When you consider the fourth industrial revolution that Dan Ies describes, TSMC isn't just a part of the narrative. It is the narrative, the golden ticket of the AI age, supplying the fundamental silicon structure for an entirely new global economy.
Real quick before we move on, if this is hitting the mark for you, drop a like and hit subscribe. It really supports my small channel. And hey, if you're thinking about retirement or investing in gold, I've got a free guide on gold IAS linked below. Stick around. I'll go over why it's useful at the end of this video.
Next up, in the major AI player market, ASML Holding Incorporated, ticker symbol ASML, the powerful Dutch firm essential for every advanced chip in existence. Without ASML, there would be no TSMC, no Nvidia, and fundamentally no AI revolution. It holds the exclusive position as the only company on the planet capable of producing EUV lithography machines.
The hyperprecise instruments used to etch the microscopic patterns onto the world's most sophisticated chips. Each of these machines costs upwards of $200 million and requires months to assemble from components sourced across Europe. ASML is the sole source, a genuine monopoly on the single most complex process in modern manufacturing.
Its recent quarterly report solidified its dominant position, achieving over 7.5 billion euros in sales, maintaining gross margins above 50% and seeing a continuous flood of new orders as chip makers like TSMC, Intel, and Samsung race to boost their production capacity. Here's why ASML's story gets even more compelling.
Even as AI algorithms become more efficient, the worldwide need for compute power is soaring. Economists call this Jieven's paradox. The more efficient something becomes, the more widely it's adopted and used. This is precisely what's happening with AI. As intelligence becomes cheaper to use, we consume it more.
And that means a relentless need for more chips, more silicon wafers, and naturally more demand for ASML's machines. Its newest EUV system, the NXE 3800E, offers 30% higher productivity than its predecessor, helping manufacturers meet the explosion in demand for AI processors. And with hypers scale giants like Microsoft and Google planning to spend record amounts in their 2025 capital expenditure budgets, ASML's gatekeeper role in the semiconductor supply chain is only getting stronger.
Even facing challenges like export restrictions and supply chain complexity, ASML's monopoly status guarantees it unmatched pricing power and stability. In this AI gold rush, ASML is the company selling the indispensable tools and every major tech player is forced to buy from them. ASML is quite literally the guardian of the digital era.
Finally, let's turn our attention to a company that is smaller, less widely known, but potentially just as revolutionary. Marvel Technology Incorporated, ticker symbol MRVL. While Nvidia commands the GPU market, Marvel is quietly pioneering a change in custom-designed AI chips. A shift that could totally remake how major tech firms power their data centers.
Marvel specializes in designing chips known as ASEX, application specific integrated circuits, chips specifically engineered for particular tasks. Rather than relying on standard one-sizefits-all GPUs, hyperscalers like Amazon and Microsoft are engaging Marvel to create chips tailorade for their proprietary AI models.
Amazon's Tranium and Inferentia processors, key parts of the AWS AI infrastructure, are built using Marvel's core technology. Similarly, Microsoft's Maya AI accelerators depend on Marvel's expertise. These specialized chips perform repetitive processes much more efficiently and affordably, lowering the dependence on Nvidia while dramatically improving performance for large-scale AI operations.
This trend marks the start of what analysts are calling the custom chip revolution. As AI infrastructure matures and becomes more specialized, companies want chips fine-tuned to their exact specifications. And that's precisely where Marvel excels. Despite being a smaller player compared to rivals like Broadcom or Nvidia, Marvel has cemented its position as the go-to engineering partner for the biggest names in tech.
From a financial perspective, the company is aggressively investing in R&D, accepting lower immediate profits to secure a long-term leadership position in this expanding niche. It boasts solid balance sheet strength, crucial partnerships with the world's leading cloud platforms, and access to a market that could grow exponentially as more businesses develop their own AI optimized silicon.
Marvel's situation is one of high risk with equally high potential, an asymmetric opportunity within the AI ecosystem. It might not get the same media attention as Nvidia or AMD, but it's playing a critical role in pushing the AI industry past generic chips toward customized architectures. In this new age of AI, Marvel isn't just keeping up.
It's actively helping to rewrite the rule book. As Dan Ies boldly stated, this is not a bubble waiting to pop. It is the fourth industrial revolution, and we are truly just getting started. The three companies we've discussed, the world's dominant chip manufacturer TSM, the unparalleled lithography gatekeeper, ASML, and the rising custom chip innovator MRVL represent the fundamental structural core of the AI economy.
They aren't simply reacting to the AI boom, they are the source of its power. The defining theme of this decade is infrastructure. The digital railways are being laid right now underneath the surface of the intelligent world we are building. Chips, data centers, and raw power are quickly becoming the new drivers of economic power.
And these specific companies are situated squarely in the center of that massive transformation. If the previous era was defined by software, this one belongs to silicon. And those who grasp where the real enduring value is being created will be the ones who ultimately shape the future. The AI super cycle has arrived and the best part of the party is still ahead of
Comments 0