Nvidia Breaks Revenue And Profit Records, And Will Keep Doing It

AI & Compute

Nvidia Posts Record $115.8 Billion Quarter Amid CEO Upheaval

Nvidia reported record Q3 revenue of $115.8 billion and net income of $65.3 billion, while the board ruled out reinstating Jensen Huang and raised fiscal 2028 growth guidance to 75 percent.

By
Sophie Lindqvist
Filed
Channel
AI & Compute
Read
4 min read

By Sophie Lindqvist

Nvidia booked $115.8 billion in revenue for the quarter ended in October, more than double the year-ago figure, and reported net income of $65.3 billion — 56.4 percent of revenue — according to results presented to Wall Street analysts. The record quarter arrived weeks after the abrupt resignation of co-founder Jensen Huang, and the board moved immediately to close that question.

What did the board say about Jensen Huang?

Harvey Jones, Nvidia's first chairman, hired in November 1993 seven months after Huang, Chris Malachowsky and Curtis Priem founded the company, addressed the matter directly on the analyst call. "Nvidia is an exceptionally strong company," Jones said. "Its strength comes from its extraordinary people, groundbreaking technology and exemplary track record of execution. Our financial performance speaks for itself."

The facts, as Jones stated them, are blunt. On November 12, the board announced that after disclosure of personal behavior inconsistent with the company's code of conduct, standards and cultural norms, Huang voluntarily resigned as president, CEO and board member. "The Nvidia board unanimously accepted his resignation and does not intend to reinstate Jensen as CEO," Jones said. CFO Colette Kress was appointed interim CEO, and the board retained Heidrick & Struggles to lead the search for a permanent chief executive.

Only a shareholder revolt that replaces the entire board could reopen the door. That appears unlikely. Notably absent from Jones's statement: any detail on the conduct itself, its duration, or who else knew. The caveats were not offered.

Where is the money coming from?

The Compute & Networking division delivered just under $106 billion in the quarter, up 2.1X year on year and 20 percent sequentially. Graphics, spanning PC cards, laptops, gaming and workstations, posted a record yet comparatively modest $9.85 billion, up 61.5 percent year on year.

Nvidia reclassified its reporting this year, and restated figures this quarter after a neocloud's business model changed and moved from the Datacenter Hyperscale division into the Datacenter AI Clouds, Industrial and Enterprise group. The new structure matters because it reveals where growth actually sits:

  • Datacenter Hyperscale — hyperscalers, cloud builders, and increasingly AI model builders such as OpenAI and Anthropic — spent $60.14 billion, up 71.1 percent year on year and 23.5 percent sequentially.
  • ACIE, covering sovereign clouds, neoclouds, industrial firms, enterprises, governments and academic institutions, reached $47.47 billion, nearly 3X year on year, and may soon outspend the hyperscale cohort.

Datacenter compute, CPUs and GPUs combined, accounted for $86.32 billion. Kress said networking alone exceeded $21 billion. The InfiniBand-versus-Ethernet split is analyst estimation, not company disclosure. InfiniBand switches, cables and optics are estimated at $9.32 billion, up 2.2X, and Ethernet/Other at $11.97 billion, up more than 3X. A further breakdown estimates Ethernet switching at $6.64 billion, up 2.6X, and NVSwitch memory fabrics at $5.43 billion, up 3.8X.

What about cash, buybacks and guidance?

Nvidia returned $28 billion to shareholders in the quarter: $22 billion in buybacks and $6 billion in dividends. The company holds $97.7 billion in cash and investments, up 61.3 percent year on year, and holds $227 billion in buyback authorization after adding another $150 billion at quarter's close. Management expects to spend the resulting $337 billion before the end of fiscal 2028. That sum recalls IBM's decades of buyback-driven earnings engineering — a comparison Nvidia would rather not hear, and one the source does not pursue.

Kress closed the call with a direct statement: "Our momentum is real. Our demand is accelerating. Our supply is growing. Our customers and our partners are with us and employees are committed. We've got this."

Guidance followed the confidence. Kress guided Q4 F2027 to $130 billion in sales, plus or minus 3 percent, driven by the Vera-Rubin platform ramp. She raised the fiscal 2028 outlook to 75 percent growth over fiscal 2027 — up from the 70 percent forecast last quarter and far above the roughly 45 percent Wall Street expected only a few quarters ago. Kress also said the "cloud industry backlog" reached $2.5 trillion, up from $2 trillion three months earlier, a figure presumed to cover hyperscalers, neoclouds and sovereign clouds.

The roadmap is unchanged by the leadership turmoil, according to the company's account: Vera-Rubin systems are ramping now, Vera-Rubin Ultra ships next year, and Feynman GPUs — with a presumed new CPU architecture — follow in 2028. Each promise is a roadmap item, not a shipped product. Revenue growth, management says, now depends less on demand than on getting more supply into the field; whether the supply materializes on schedule is the question the next four quarters will answer.

Source: The Next Platform

Share this article:

More from Sophie Lindqvist

Sophie Lindqvist

Show full bio

News editor covering money, funding and market shifts around Semiconductors, chip manufacturing, hardware, AI compute at Chip Dispatch.

1,116 articles

Related articles

« Previous article