Nvidia's $5 Trillion Milestone: How a New AI Model and a Speck of Silicon Are Reshaping the Landscape

Nvidia crosses $5 trillion with new AI software, while Micron surges 560% as analysts predict a structural memory supercycle driven by AI demand.

Nvidia has crossed the $5 trillion market-cap threshold, a feat that would have seemed improbable even a year ago. The stock has surged roughly 27% in the past month alone, hitting a fresh all-time high of €182.28. But the catalyst behind this latest leg higher isn't just another GPU—it's a software play that could tighten the company's grip on the entire AI ecosystem.

The chipmaker unveiled Nemotron 3 Nano Omni, a multimodal open-source model that processes images, speech, and text in a single system. With 30 billion parameters but only 3 billion activated per inference thanks to a Mixture-of-Experts architecture, the model runs on a single GPU. Nvidia claims a ninefold throughput advantage over comparable open models. Early adopters include Foxconn, Palantir, and Oracle.

The strategic message is unmistakable: Nvidia's models run best on Nvidia's hardware, and Nvidia's hardware is optimized for Nvidia's models. It's a closed loop that directly challenges the model-plus-cloud offerings from Google, Amazon, and Microsoft.

The financials back up the narrative. In the fourth quarter of fiscal 2026, revenue climbed 73% to $68 billion, while adjusted earnings per share rose 82% to $1.62. Management guided for around $78 billion in the current quarter—comfortably above consensus estimates. Bank of America reiterated its buy rating with a $300 price target. The next major catalyst arrives on May 20, when Nvidia reports first-quarter results for fiscal 2027.

Micron's $1,000 Bet on a Structural Memory Supercycle

While Nvidia dominates headlines, Micron is quietly writing its own remarkable story. The stock has gained more than 560% over the past twelve months and hit a new closing record, trading at €449.30—up roughly 4% on the day.

DA Davidson initiated coverage with a buy rating and a $1,000 price target, the most bullish call in the semiconductor sector this year. The thesis: artificial intelligence is generating an unusually long memory cycle. Each new wave of computing power pulls more demand behind it, and the market is structurally underestimating Micron's share of that loop. TD Cowen followed by raising its target from $550 to $660.

The numbers support the optimism. In the first fiscal quarter of 2026, Micron posted adjusted earnings of $4.78 per share—well above the $3.94 consensus—on revenue of $13.64 billion, up 57% year over year. Cloud storage revenue nearly doubled. For the second quarter, management guided for $18.7 billion in revenue and adjusted EPS of $8.42, with order books extending into 2027.

The entire HBM production for calendar 2026 is already contractually committed, including the next-generation HBM4. The addressable market is expected to grow from $35 billion in 2025 to roughly $100 billion by 2028. To keep pace, Micron plans capital expenditures of about $20 billion this fiscal year. Of 43 analysts covering the stock, 38 rate it a buy or strong buy. The forward P/E of 9x looks remarkably cheap given the growth profile.

ASML: Building Buses Faster Than Anyone Can Drive Them

The Dutch lithography specialist is cranking up production. This year, ASML will deliver at least 60 EUV systems—36% more than the 48 machines shipped last year. The next expansion phase targets 80 units annually. To sustain this pace, the company is investing €2.2 billion in infrastructure, with new cleanrooms already operational in the US, Germany, and South Korea.

First-quarter results underscored the momentum: €8.8 billion in revenue, a 53% gross margin, and €2.8 billion in net profit. The full-year guidance was raised to €36–€40 billion. RBC Capital responded by lifting its price target from $1,625 to $1,700.

Beyond the core business, a €1.3 billion strategic partnership with French AI startup Mistral AI caught the market's attention. One concrete example: a wafer-defect diagnostic process that previously took over ten hours now shrinks to eight minutes using a Mistral model.

The China revenue share is declining as planned—from 41% in 2024 to 33% last year and an expected 20% this year. CFO Roger Dassen argues that in an undersupplied market, restrictions in one region are compensated by higher investments elsewhere. ASML trades at around €1,185, roughly 8.5% below its February high of €1,295. The export-control debate could shape the valuation more than any operating metric in the coming months.

TSMC: Five 2nm Factories Running Simultaneously

At its North America Technology Symposium, TSMC made clear that restraint is no longer an option. Five factories for the 2-nanometer process will enter volume production this year—two in Hsinchu and three in Kaohsiung. Such aggressive expansion is unprecedented in the company's history.

Between 2026 and 2028, TSMC plans annual capacity growth of 70% for 2nm chips. By year-end, monthly capacity should reach 100,000 wafers. Meanwhile, 3nm production is expanding from 150,000 to 180,000 wafers per month. The entire capacity of the first two 2nm factories is already spoken for: Apple secured more than half, with Qualcomm, MediaTek, AMD, and Nvidia taking the rest. Yield rates for the 2nm process are outperforming the 3nm predecessor, despite the more complex nanosheet architecture.

First-quarter revenue rose 39.2% year over year. 98% of analysts rate the stock a buy. Already in development is the next node, A16, which introduces backside power delivery to meet rising demands from AI and automotive applications.

Intel: The Comeback That Nobody Saw Coming

Six consecutive quarters of beating both internal guidance and analyst expectations—the latest report on April 23 was the most impressive yet. Revenue hit $13.6 billion, up 7% year over year and $1.4 billion above consensus. Adjusted EPS of $0.29 crushed the $0.01 forecast.

The driver: AI-related business accounted for 60% of revenue, growing 40% year over year. Data center revenue climbed 22% to $5.1 billion. The CPU market, long considered dormant, is benefiting from agent-based AI applications shifting computing demand beyond GPUs and back to central processors.

For the second quarter, Intel guided for revenue of $13.8–$14.8 billion—again well above market expectations of $13.07 billion. Strategically, the announcement that Intel will join Elon Musk's Terafab chip complex in Austin, Texas, to develop high-performance chips for SpaceX, xAI, and Tesla turned heads. Freedom Broker upgraded the stock from hold to buy, quadrupling its price target from $25 to $100. The stock trades at €73.07, having more than doubled in a month.

Not everyone is buying the narrative. Yield issues with the 18A process and the challenge of poaching long-time TSMC customers remain real structural hurdles. Some observers see Intel less as a strategic winner than as a company that stumbled into a hot trend.

The Sector Dynamics at a Glance

• Semiconductor industry 2026: Total revenue on track toward $975 billion, with nearly $500 billion coming from generative AI chips alone
• Memory market: Nearly all available production capacity for 2026 is already reserved
• Lithography: ASML physically cannot build machines fast enough to meet demand
• Foundry: TSMC's 2nm capacity for the first factories is fully booked
• Nasdaq: Up 15% in April—the best month since April 2020

Between Record Runs and Capacity Constraints

The industry is showing an unusually broad rally. All five heavyweights are moving in the same direction—each for different reasons. TSMC is the foundry running five factories simultaneously. ASML supplies the machines everyone needs but nobody can get fast enough. Micron serves the sold-out memory demand. Nvidia is building the software-plus-hardware ecosystem that serves everyone. And Intel is capturing a slice of AI demand through CPUs that nobody had on their radar.

Near-term catalysts are clear: Nvidia's earnings on May 20, Micron's numbers in mid-June, TSMC's ongoing yield data on the 2nm ramp. For Intel, the question is whether the CPU comeback thesis extends beyond the current agentic AI wave—management forecasts strong server CPU demand through 2027. Confirmation would validate the entire turnaround story. For ASML, the political dimension looms large: the export-control debate could shape the valuation more than any operating metric.

Supply chains are running at the limit. That's exactly what's driving the stock prices—and exactly what poses the biggest risk if demand cools even slightly.