Nvidia's Record Numbers and $118 Billion Payout Mask a Harder Task: Proving Industrial AI Can Scale

Nvidia reports $81.6B revenue, beats EPS, splits data center into hyperscale and ACIE, announces Kawasaki robotics deal, and authorizes $80B buyback amid valuation concerns.

Nvidia delivered the kind of quarter most companies only dream of — $81.6 billion in revenue, earnings per share that beat Wall Street expectations, and a fresh $80 billion buyback authorization. Yet the stock drifted lower. The tension between operational brilliance and valuation discipline is now the defining feature of Nvidia's narrative, but a deeper strategic pivot may be unfolding beneath the surface.

The chip giant is reshaping how it reports its business, splitting the dominant data-center segment into two parts: Hyperscale, which generated $37.9 billion in the fiscal first quarter, and ACIE — AI Clouds, Industrial and Enterprise — which contributed $37.4 billion. That near parity is no accident. The new structure reveals a structural shift: industrial and enterprise demand for computing power is now running almost neck-and-neck with the hyperscaler juggernaut. A third segment, Edge Computing, pulled in $6.4 billion, up 29% year over year, bundling gaming, professional visualization, and robotics.

That edge segment just got a powerful endorsement. Nvidia announced a partnership with Kawasaki Heavy Industries, the Japanese industrial giant, to integrate its robotics and agentic AI technologies into manufacturing lines. CEO Jensen Huang had telegraphed the move with his declaration that "agentic AI has arrived" — systems capable of executing complex, multi-step tasks autonomously. The deal lands squarely in the new Edge Computing reporting line and signals where Nvidia is expanding beyond the data center: into autonomous industrial robotics, physical production environments, and vehicle systems.

On the capital returns front, the board authorized a $80 billion expansion of the stock buyback program, adding to the $38.5 billion that remained from prior approvals for a total capacity of $118.5 billion. The quarterly dividend was hiked from $0.01 to $0.25 per share, payable June 26 to shareholders of record June 4. In the just-completed quarter, roughly $20 billion flowed back to shareholders through buybacks and dividends. The free cash flow generating that firepower stood at $49 billion.

The numbers behind those payouts remain extraordinary. Revenue for the first quarter of fiscal 2027 rose 85% from a year ago and 20% sequentially. The data-center segment alone hit $75.2 billion, a 92% year-over-year increase. On a GAAP basis, diluted earnings per share came in at $2.39; on a non-GAAP basis, the figure was $1.87, topping the consensus estimate of $1.78. For the current quarter, Nvidia guided revenue of $91 billion, plus or minus 2%, with the explicit caveat that the forecast excludes data-center compute revenue from China due to ongoing trade restrictions.

The guidance also reflects confidence in the upcoming Vera Rubin platform. First shipments are slated for the third quarter of 2026, with volume ramp in the fourth. Huang expects demand to outstrip supply across the product's entire life cycle. Gross margins are expected to remain elevated: 74.9% on a GAAP basis and 75.0% non-GAAP, each with a 50-basis-point cushion.

Despite all this, the stock closed Friday at €185.46 in European trading, down 1.83% on the day and 4.34% for the week — roughly 8% below its 52-week high from May 14. At a market capitalization of about $5.4 trillion and a price-to-earnings multiple of roughly 30.5 times estimated calendar 2026 earnings, the execution bar is high. Technical indicators are mixed: the relative strength index sits at 40.5, while the shares trade 10.1% above the 50-day moving average and 16.0% above the 200-day line.

Two key events loom. On May 28, Nvidia will present at the TD Cowen Technology, Media & Telecom Conference, where investors will scrutinize comments on demand trends and the China gap. That same day, the US releases April PCE inflation data, a metric that could shift interest rate expectations and directly impact richly valued tech stocks. On June 4, management appears at the BofA Global Technology Conference. Both sessions will be webcast and archived for 90 days.

Nvidia now operates in two distinct arenas. One is the hyperscale AI infrastructure business that continues to print record numbers. The other is an emerging industrial and physical AI frontier — embodied by the Kawasaki deal — that the company is trying to turn into a growth engine. The market, for now, wants proof that both can deliver simultaneously.