Nvidia's Record Cash Hoard Powers a Three-Continent Expansion as the Market Wavers

Nvidia's Q1 revenue surged 85% to $81.6B, beating expectations, but the stock remained flat amid rising bond yields and high valuations. The company unveiled a new reporting structure, $80B buyback, and global infrastructure moves.

Nvidia delivered a quarter that most companies would trumpet from the rooftops: revenue of $81.6 billion, up 85%, and earnings per share of $1.87 that handily beat the $1.75 consensus. Yet the stock barely budged, closing at €183.12 in Europe on Thursday — essentially flat on the day and down 3.07% for the week. The 12-month gain remains punchy at 53.29%, but the shares are now 8.92% below their mid-May peak, and the RSI of 36.4 suggests momentum has cooled.

The disconnect reflects a market that already priced in the fireworks. Nvidia's guidance for the current quarter — around $91 billion in revenue, well above the $87.2 billion consensus — impressed, but the specter of rising bond yields (the 30-year U.S. Treasury hit 5.197% recently) and lofty valuations kept buyers cautious. The annualized volatility of 37.05% underscores the tug-of-war between operational acceleration and investor wariness.

A Divvy, a Buyback, and a $48.6 Billion Cash Gush

Underneath the stock's sideways drift, Nvidia's financial muscle is on full display. The board authorized $80 billion in fresh share repurchases, adding to $38.5 billion still available from the prior program. The quarterly dividend — hiked from a token $0.01 to $0.25 per share — will be paid on June 26 to holders of record on June 4. Free cash flow hit $48.6 billion, and the gross margin remained thick at 75%.

The data-center segment, which contributes $75.2 billion in revenue, grew 92% year-over-year and remains the engine room. But Nvidia is now reshaping how it tells that story.

A Reporting Overhaul to Sharpen the Narrative

On May 28, at the TD Cowen TMT conference, Nvidia unveiled a new reporting structure. Two platforms will now take center stage: Data Center and Edge Computing. The move is more than cosmetic. Inside Data Center, the company will break out customer groups — Hyperscale and ACIE (AI Clouds, Industry, and Enterprise) — giving investors a clearer lens on state-run AI programs and corporate clients. Edge Computing bundles devices for physical and agentic AI, including PCs, robotics, automotive components, and AI-RAN base stations.

The message is that Nvidia's growth is broadening beyond hyperscalers. CEO Jensen Huang has described demand for AI infrastructure as "parabolic," driven especially by agentic AI applications. The Vera CPU project, targeting a $200 billion market, could add up to $20 billion in revenue this fiscal year alone, according to some analysts. Tigress Financial pegs the stock at $425; the average analyst target is $305.38.

A Global Footprint That Reads Like a Geopolitical Chessboard

While the numbers grab headlines, Nvidia is quietly building infrastructure on three continents. The most symbolic move: Huang has joined the advisory board of the School of Economics and Management at Tsinghua University in Beijing — a body chaired by Apple's Tim Cook and including Elon Musk, Satya Nadella, Mark Zuckerberg, and Jamie Dimon. For a CEO whose company is barred from selling its top-tier H100 and A100 AI accelerators to Chinese customers under U.S. export controls, the seat is a diplomatic channel as much as an academic honor. Huang has long warned that a full pullback from China would cede ground to local rivals Huawei, Cambricon, and SMIC. Nvidia's modified chips — H20 and B30A — try to stay within U.S. rules while still serving Chinese demand. Currently, the company reports no data-center compute revenue from China.

Israel is another prong. Nvidia has signed a 10-year lease for an 11-story building in Yokneam Ofer Park, adding roughly 29,000 square meters at a cost of about 230 million shekels (index-linked). That facility is set for completion in 2028 and will bring Nvidia's total footprint in the park to 67,000 square meters. A separate planned headquarters in Kiryat Tivon will span 160,000 square meters, underscoring a long-term bet on Israeli R&D talent.

Then there's Taiwan. Huang stressed that Nvidia now spends roughly $150 billion annually with its manufacturing partners there — about ten times the level of five years ago. That cash flows to partners who build products and manage the supply chain. Without that industrial base, the AI boom would struggle to translate into shippable hardware.

The Road Ahead: Vera Rubin and the Next Test

Nvidia's next architecture, Vera Rubin, is slated for 2026, and early demand signals are reportedly even stronger than the current Blackwell cycle. For the near term, all eyes are on the Q2 delivery: can the company hit that $91 billion target while keeping data-center margins steady? If so, the China diplomacy and global expansion get more breathing room. If growth slows, the specter of hyperscalers developing their own chips — Google, Microsoft, and others are already doing it — will move to the forefront.

For now, Nvidia is playing a multi-dimensional game: record cash flow, an $80 billion repackaging of capital, a refined financial narrative, and a physical presence that spans Beijing boardrooms, Israeli tech parks, and Taiwanese fabs. The stock may be treading water, but the company is building foundations across multiple time zones.