NVIDIA Organizational Structure

About the company

NVIDIA designs accelerated-computing processors, the systems and networking that turn them into data-centre infrastructure, and the CUDA software stack most AI work runs on. It was incorporated in California in April 1993 by Jensen Huang, Chris Malachowsky and Curtis Priem, reincorporated in Delaware in 1998, and is run from Santa Clara. Revenue for fiscal 2026, the year ended 25 January 2026, was 215.9 billion dollars, 65 percent higher than the year before.

The organisation behind that number is small and unusually technical. Roughly 42,000 people work across 38 countries, and about 31,000 of them sit in research and development; fewer than 11,000 cover sales, marketing, operations and administration for the entire company. More than 80 percent hold technical roles and over half hold an advanced degree.

The structure is worth studying because it did not change when the revenue did. A company that multiplied several times over in three years still runs as one functional organisation with no divisional general managers, and its chief executive has said publicly that around 60 people report to him. The conventional answer to that kind of growth is more layers. NVIDIA's answer was more width.

NVIDIA Corporation organizational chart

NVIDIA Corporation's reporting structure as of 2026-08. Drag to pan, scroll to zoom — or open it in QueryChart and edit it as your own.

One functional company, and a CEO span measured in dozens

The chart depicts NVIDIA Corporation itself. There is no holding company above it, no separately governed operating subsidiary beneath it, and no divisional chief executive anywhere between. Everything below Jensen Huang is either a company-wide function — engineering, research, operations, field operations, finance, legal, people — or a business and industry area led by a vice president who owns a market rather than a profit and loss.

The two segments NVIDIA reports, Compute & Networking and Graphics, are a financial-reporting construct rather than the management structure. The company names its chief executive as the chief operating decision maker and discloses that expenses attributable to each segment are not regularly provided to him: costs are reviewed on a consolidated basis and shared costs are allocated across both segments because they rest on one unified architecture. Nobody in the chart runs a segment.

The formally designated officer set is tiny for the size of the business. The fiscal-2026 10-K names five Section 16 executive officers: the chief executive, the chief financial officer, and executive vice presidents for worldwide field operations, operations, and legal. There is no chief operating officer, no president other than Huang, no chief product officer and no chief marketing officer. All go-to-market activity sits in a single Worldwide Field Operations organisation rather than in regional divisions, and the industry verticals — healthcare, automotive, telecommunications, robotics — are led by product vice presidents who sit alongside the engineering leaders rather than inside the sales organisation.

Two features of the chart carry a date stamp. NVIDIA Research is held apart from the product engineering lines as a roughly 200-scientist organisation under Chief Scientist and SVP of Research Bill Dally, with an applied-research arm whose job is to carry results across into shipping products. And two senior seats moved in 2026: Scott Gawel succeeded Donald Robertson as chief accounting officer in May, and Ajay Puri announced his retirement in June after 21 years running worldwide field operations, with Nicholas Parker joining from Microsoft on an anticipated start date of 24 August 2026 while Puri stays on as a senior adviser. The chart shows the incoming arrangement. Below the five named executive officers, NVIDIA publishes no reporting lines at all, so the layer beneath them is reconstructed from the company's own public titles.

Key leadership roles

1. Jensen Huang

   President and Chief Executive Officer — Co-founder, in the role since 1993; has said publicly that roughly 60 leaders report to him and that he holds no one-on-one meetings with any of them.

2. Colette M. Kress

   Executive Vice President and Chief Financial Officer — One of the five Section 16 executive officers named in the fiscal-2026 10-K.

3. Nicholas Parker

   Executive Vice President, Worldwide Field Operations — Joining from Microsoft on an anticipated start date of 24 August 2026, succeeding Ajay Puri, who held the seat for 21 years and remains as a senior adviser.

4. Debora Shoquist

   Executive Vice President, Operations — Owns manufacturing, supply and quality across every product line; NVIDIA has no factories of its own.

5. Timothy S. Teter

   Executive Vice President, General Counsel and Secretary — The fifth of the five designated executive officers, and the only legal seat on the chart.

6. Michael Kagan

   Chief Technology Officer — Came in with the Mellanox acquisition; NVIDIA does not publish his reporting line, so the line to the CEO here is inferred from the flat structure.

7. Bill Dally

   Chief Scientist and Senior Vice President of Research — Runs NVIDIA Research, roughly 200 scientists deliberately held outside the product engineering lines.

8. Jonah Alben

   Senior Vice President, GPU Engineering — One of the longest-tenured engineering leaders on the chart.

9. Dwight Diercks

   Senior Vice President, Software Engineering — Owns the software stack that keeps every market on one architecture.

10. Kevin Deierling

    Senior Vice President, Networking — NVIDIA has publicly described two senior vice presidents of networking; only the one named on its own conference listings is shown, and the split between them is not public.

11. Chris A. Malachowsky

    Founder and NVIDIA Fellow — A co-founder in a technical fellow role rather than a line-management one, so the box beneath him is deliberately empty.

12. Kristin Major

    Senior Vice President, Human Resources — Joined from Hewlett Packard Enterprise; owns the people function for a workforce that is roughly three-quarters R&D.

13. Scott Gawel

    Vice President and Chief Accounting Officer — Appointed in 2026 on Donald Robertson's retirement; reports to the chief financial officer rather than to the CEO.

14. Kari Briski

    Vice President, AI Software Product Management — Her title is NVIDIA's own, but no source states her reporting line — placing her under software engineering is an inference, not a documented line.

How this structure supports the business

One functional organisation over one architecture is what lets NVIDIA amortise the same silicon, the same CUDA stack and the same networking investment across gaming, data centre, automotive, robotics and telecommunications. A divisional company would build a version of that stack per division; here the engineering leaders own their discipline for every market at once, and the industry vice presidents buy from them rather than duplicating them.

The extreme span is an information design as much as an org design. Huang's stated practice of holding no one-on-ones and raising problems in front of the whole leadership group means the top of the company hears the same thing at the same time, without a management layer deciding what is worth passing up. That is what makes it possible to redirect thousands of engineers within a product cycle rather than a planning cycle.

Consolidating every commercial relationship into one Worldwide Field Operations organisation matches how the largest customers actually buy. A hyperscaler or a sovereign AI programme purchases a system — silicon, networking, systems and software together — and a divisional sales structure would present that customer with several negotiating counterparties for one purchase.

Advantages and disadvantages

One architecture serving every market means a single engineering investment is recovered across gaming, data centre, automotive and robotics at once.

No business unit owns an end-to-end P&L, so the real profitability of a smaller line such as automotive or professional visualisation is close to invisible — to management and to investors alike — and a weak line is correspondingly hard to identify or cut.

A span of roughly 60 direct reports with no filtering layer keeps the decision with the one person who can see the whole company, and moves resources at product speed.

It concentrates execution risk in one individual, offers no obvious internal succession path, and makes the chief executive the scheduling bottleneck for decisions a divisional company would settle a level or two down.

A single Worldwide Field Operations organisation gives every large customer one commercial interface to a company whose products they buy as a system.

The entire revenue relationship sits in one seat, so a handover of that seat — Puri to Parker in August 2026 — touches every customer at once, with no divisional sales chiefs to absorb it.

Holding NVIDIA Research separate under a chief scientist protects long-horizon work from quarterly product pressure, and several product lines started there.

Research outside the engineering lines creates a permanent transfer problem, which is why a distinct applied-research group has to exist purely to carry work across the boundary — overhead an integrated R&D function would not need.

Interesting facts

• Roughly 60 leaders report directly to the chief executive, and he holds no one-on-one meetings with any of them; problems are presented to the whole leadership group so information is not filtered by a management layer.
• For a company with 215.9 billion dollars of fiscal-2026 revenue, NVIDIA designates only five Section 16 executive officers — CEO, CFO, and EVPs for worldwide field operations, operations and legal. There is no COO, no divisional president, no chief product officer and no CMO.
• The chief executive is formally the chief operating decision maker, and NVIDIA discloses that expenses attributable to each reportable segment are not regularly provided to him, which makes segment profitability a reporting artefact rather than a management unit.
• About 31,000 of roughly 42,000 employees work in research and development, so the technical organisation is close to three times the size of everything commercial and administrative combined.
• Turnover was 3.7 percent in fiscal 2026 and more than 40 percent of that year's new hires came from employee referrals, meaning the organisation largely grows through its own network rather than through the open market.
• In December 2025 NVIDIA expanded the organisation by licensing Groq's language-processing-unit technology and hiring its staff for roughly 17 billion dollars, booking 14.4 billion dollars of goodwill attributed primarily to the workforce — an org-building move structured as a licence-plus-hire rather than an acquisition.

Frequently asked questions

What type of organizational structure does NVIDIA use?

A functional, unitary structure. Engineering, research, operations, worldwide field operations, finance, legal and people are company-wide functions serving every product line, and the business and industry areas are led by vice presidents who own a market rather than a profit and loss. No executive below the chief executive runs a division.

How many people report directly to Jensen Huang?

NVIDIA does not disclose the number. Huang has said publicly that it is roughly 60; earlier reporting put it at 55, and one 2026 analysis counted 36. What is consistent across all of them is the order of magnitude — a span an order larger than a typical large-cap chief executive, with no one-on-one meetings and no chief operating officer in between.

Does NVIDIA have divisions or business units?

Not as management units. It reports two segments, Compute & Networking and Graphics, but states that its chief operating decision maker is not regularly provided with expenses attributable to either, and that shared costs are allocated across both because they rest on one architecture. The segments are a disclosure boundary, not an operating one.

Who are NVIDIA's executive officers?

The fiscal-2026 Form 10-K names five: Jensen Huang as president and chief executive officer, Colette M. Kress as chief financial officer, and executive vice presidents for worldwide field operations, operations and legal. The worldwide field operations seat is in transition, with Nicholas Parker joining on an anticipated start date of 24 August 2026 as Ajay Puri retires from the role.

Can I edit this NVIDIA org chart?

Yes. Open it as your own chart and it becomes an ordinary QueryChart org chart — rename the roles, add or remove reports, and edit it as a table or on the canvas. It is a useful starting point if you are testing what a very wide, function-only structure would look like at your own headcount.

Use this org chart template

Open NVIDIA's structure as an editable org chart and use it to test what a functional organisation with a very wide executive span looks like at your own size.

Open this org chart as your own