Q1 revenue $81.6bn (+85% YoY), data centre $75.2bn, net income $58.3bn. The quarterly dividend moves from a penny to 25 cents per share. The buyback is the second $80bn authorisation in three quarters.
Nvidia reported first-quarter revenue of $81.6bn on Wednesday, up 85% year-on-year and 20% from the prior quarter, and guided second-quarter revenue to about $91bn (plus or minus 2%) against an analyst consensus of $86.84bn.
The board separately authorised an additional $80bn of share buybacks and lifted the quarterly dividend from one cent to 25 cents per share. The stock moved up on the print before settling roughly flat in extended trading.
Data-centre revenue inside the quarter came in at $75.2bn, against the average analyst estimate of $72.8bn, as the dominant beat. Net income for the February-April period was $58.3bn, up 37% sequentially and more than 200% year-on-year, on the run rate Al Jazeera’s coverage framed as ‘record profit and revenue amid the AI chip boom’.
The buyback authorisation is the second $80bn programme the board has approved inside three quarters. Combined, the company has authorised over $160bn of repurchases against a market capitalisation that has, on the May 20 close, traded above $4tn for most of the past quarter.
The quarterly-dividend bump, from one cent to 25 cents, is mathematically immaterial against the buyback envelope but signals a balance-sheet posture that has moved decisively from growth-financing to capital-return on the discretionary-cash margin.
The Q2 guide implies sequential growth of about 12% on the Q1 base. The Street’s read is that the guide is conservative against the publicly visible hyperscaler-capex commitments for the second half of the year. AWS, Microsoft, Google and Meta have collectively guided to roughly $470bn of 2026 capex, the majority of which clears through Nvidia silicon.
The $115bn-plus Meta capex line, the AWS GB200/GB300 NVL72 deployments and the Google-Blackstone $25bn TPU-cloud joint venture are the visible demand-side scaffolding for the Q2-Q3 print.
On the competitive-context line, Nvidia’s data-centre share against the AI-accelerator alternative cohort remains the structural story. Tenstorrent’s takeover conversations with Intel and Qualcomm and Alibaba’s T-Head Zhenwu M890 announcement represent the two visible non-Nvidia compute paths, the US-side RISC-V/x86 alternative and the China-side domestic accelerator track respectively.
Neither has yet shipped at the volume that would dent the data-centre growth line. The Trump-Xi licensing track on H200 sales to Chinese customers remains the swing variable for the FY27 second-half guide.
Nvidia did not disclose the time horizon over which the new $80bn buyback authorisation is expected to be executed, the geographic breakdown of the Q1 data-centre revenue (US-versus-rest-of-world), the specific Q2 GB300 NVL72 shipment volumes baked into the guide, or the H200-licensing revenue inside the China line.
Chief executive Jensen Huang’s prepared remarks on the call framed the AI infrastructure build-out as ‘still in the early innings’, the company’s standard formulation.
The next visible proof point will be the FY27 Q2 print, with the Q3 guide that will accompany it the more closely watched data point on whether the second-half capex commitments translate cleanly into Nvidia revenue.
Get the TNW newsletter
Get the most important tech news in your inbox each week.