---
title: Constant Direct Communication-Why Nvidia’s CEO Drops Everything for His Team
description: "Nvidia CEO Jensen Huang manages 36 direct reports with no 1:1 meetings. Inside the flat structure driving Nvidia's speed and its governance risks."
author: Darie Nani (Editor-in-Chief)
date: 2025-10-16T04:34:00.000Z
updated: 2026-05-16T16:16:09.635Z
canonical: https://www.sovereignmagazine.com/article/constant-direct-communication-why-nvidia-s-ceo-drops-everything-for-his-team
image: https://cdn.nanimediahouse.com/1daee9aa-58a7-48de-a68b-47610c5d556b-scaled.jpg
categories: Business Savvy
content_type: Profile
region: United States
publication: Sovereign Magazine
about:
  - type: Organization
    name: Nvidia
---

Jensen Huang has 36 direct reports. A year ago, he had 55. At the DealBook Summit in 2023, he declared: ‘The more direct reports the CEO has, the less layers are in the company.’ The [Business Insider org chart published on 14 October](https://www.businessinsider.com/nvidia-org-chart-leaders-report-to-ceo-jensen-huang-2025-10) shows who sits in Nvidia’s inner circle – and why this flat structure creates both speed and serious risk.

## What the List Actually Contains

Twenty-eight of the 36 direct reports are engineering or product leaders. That’s 78 per cent technical talent reporting straight to the CEO. The remaining eight handle business functions: finance, legal, HR and communications.

This ratio reveals Nvidia’s engineering-first DNA. While most Fortune 500 CEOs surround themselves with business executives and advisers, Huang keeps GPU architects, networking chiefs and AI platform builders at his side. Three names capture this approach:

**Ian Buck** – [CUDA’s inventor](https://blogs.nvidia.com/blog/author/ian-buck/), who joined in 2004 after his Stanford PhD. Buck turned GPU computing from graphics-only into the general-purpose platform powering today’s AI revolution. He now runs Nvidia’s multi-billion-dollar datacentre business, which is central to the shift towards [AI factories reshaping computing infrastructure](https://www.sovereignmagazine.com/article/ai-factories-are-the-new-data-centres).

**Bill Dally** – [Chief scientist since 2009](https://research.nvidia.com/person/william-dally) with serious credentials: 12 years at Stanford including department chair, over 160 patents and membership in the National Academy of Engineering. Under him, Nvidia Research has grown to over 300 scientists globally.

**Michael Kagan** – [The CTO who co-founded Mellanox in 1999](https://spectrum.ieee.org/nvidia-cto-michael-kagan-profile/about-michael-kagan) and joined Nvidia when it bought the networking company in 2020. His vision of integrating high-performance networking with GPU computing drives Nvidia’s AI infrastructure play.

## How Huang Manages 36 People

Huang doesn’t schedule regular one-on-ones. [When someone requests a meeting, he’ll ‘drop everything’](https://www.cnbc.com/2024/06/11/nvidia-ceo-i-never-schedule-one-on-one-meetings-unless-someone-asks.html) to make it happen. Even stranger: at Stanford in March, he revealed that [many direct reports earn the same amount](https://www.gsb.stanford.edu/insights/jensen-huang-how-use-first-principles-thinking-drive-decisions) – no hierarchy by pay grade.

Information moves through constant, informal communication. Huang sends frequent emails to all staff and [expects leaders to share horizontally](https://www.sovereignmagazine.com/article/nvidia-valuation-surpasses-germany-gdp-bubble) rather than hoard in silos. No telephone game through management layers because there aren’t any.

## The Upside – Speed and Product Focus

When the chip designer sits with the driver engineer and the software architect, handoffs happen in real-time. Key product owners – from DGX cloud systems to GeForce gaming to Omniverse simulation – all have direct CEO access. Traditional semiconductor companies bury product managers five levels down from strategic decisions, as seen with challenges facing companies like [Intel in the evolving AI chip landscape](https://www.sovereignmagazine.com/article/challenged-by-openai-advancement-intel-faces-a-sharp-decline-in-its-stock-price).

Huang isn’t alone in running flat. [Mark Zuckerberg manages Meta through 25 to 30 lieutenants](https://www.businessinsider.com/meta-ceo-mark-zuckerberg-management-style-direct-reports-delegation-2025-5), though not all report directly. [Elon Musk had 19 direct reports at Tesla in August](https://www.businessinsider.com/tesla-org-chart-executives-elon-musk-2025-9), down from 35 after layoffs. Huang’s 36 puts him at the extreme end.

## The Costs – Bandwidth and Governance

Human attention doesn’t scale. Even 30 minutes weekly with each direct report means 18 hours of one-on-ones alone. Add board meetings, customer visits and product reviews – the maths gets brutal.

The ‘drop everything’ approach works when teams align and the company wins. But what happens when groups clash over resources? When technical decisions need arbitration? Issues often fester in flat organisations until they explode, particularly as [competition intensifies in the AI chip wars](https://www.sovereignmagazine.com/article/the-ai-chip-wars-heat-up-how-openai-and-broadcom-are-reshaping-manufacturing-s-future).

The drop from 55 to 36 direct reports suggests Nvidia hit these limits. Where did those 19 leaders go? Without knowing whether they got promoted, consolidated or left, it’s hard to judge if this represents healthy evolution or concerning churn.

## Stakes for Nvidia’s Future

How Huang organises senior leadership directly impacts Nvidia’s ability to execute [AI ambitions](https://www.sovereignmagazine.com/article/nvidia-british-startup-ineffable-intelligence-reinforcement-learning). The direct reports map onto critical growth areas: hyperscale computing (Buck), DGX cloud (Alexis Bjorlin), enterprise AI (Justin Boitano), networking (Kagan) and Omniverse (Rev Lebaredian). Each owns a piece of the AI puzzle, even as competitors like [China’s DeepSeek challenge US tech giants](https://www.sovereignmagazine.com/article/china-s-deepseek-takes-on-us-tech-giants-what-this-means-for-project-stargate) with innovative approaches.

This creates strength and fragility. Technical leaders move fast without bureaucracy. But so much depends on Huang’s bandwidth and these specific individuals staying put. Lose one or two key people, and entire product areas stumble.

[The succession question looms largest](https://www.sovereignmagazine.com/article/john-ternus-apple-next-ceo-first-engineer-since-steve-jobs). Traditional structures create clear deputies and heirs. Huang’s flat approach makes it harder to identify who could replace him. When 36 people report to the CEO, which one is being groomed?

For acquisitions, the structure creates integration challenges. Mellanox worked partly because Kagan got direct Huang access. But can Nvidia absorb more companies without overwhelming the CEO’s capacity, especially amid geopolitical tensions like [China’s ban on Nvidia chips](https://www.sovereignmagazine.com/article/china-bans-nvidia-chips-after-decade-of-us-tech-restrictions)?

Watch for promotions, reorganisations or departures among GPU engineering, networking, enterprise AI and Omniverse leadership over the next year. Those moves signal whether Huang’s radical flatness scales with Nvidia’s ambitions – or whether even the most horizontal company eventually needs hierarchy. The names on this list aren’t just org chart entries. They’re proxies for [Nvidia’s next wave of products](https://www.sovereignmagazine.com/article/golden-handcuffs-at-nvidia-broadcom-and-amd-how-rsus-are-changing-managers-teams-and-hr-metri) and slices of risk.

[Nvidia’s stellar earnings](https://www.sovereignmagazine.com/article/why-nvidia-s-stellar-earnings-couldn-t-stop-the-tech-rout) show just how sentiment-driven tech markets can be, even for the industry leader.