How Nvidia Makes Money from AI Chips

Nvidia Corporation, incorporated in Delaware and headquartered in Santa Clara, California, sells graphics processing units, networking hardware and software to data centres, gaming consumers, professional visualisation users and automotive manufacturers, with the bulk of its revenue now generated by AI computing infrastructure.โ€ฆ

TPC Editorial Desk

By

TPC Editorial Desk

Published

20 Nov 2026

Read

7 min

How Nvidia Makes Money from AI Chips

Photo: panumas nikhomkhai / Pexels

As of 20 November 2026. This profile reports Nvidia's published business information, attributed to the sources named. It is general information, not investment, legal or tax advice, and The Platinum Capital has no commercial relationship with Nvidia unless stated. We update or correct it when the facts change.

What the Company Does

In its Form 10-K for the fiscal year ended 26 January 2025, filed with the United States Securities and Exchange Commission, Nvidia described itself as "a full-stack computing infrastructure company with data-center-scale offerings." The company stated that its full stack includes the CUDA programming model, hundreds of domain-specific software libraries, software development kits and application programming interfaces. These tools accelerate workloads such as AI model training and inference, data analytics, scientific computing and 3D graphics across industries including healthcare, telecoms, automotive and manufacturing. (Nvidia Corporation, Form 10-K filed with the U.S. Securities and Exchange Commission, 26 February 2025)

The company described a platform strategy that "brings together hardware, systems, software, algorithms, libraries, and services." In its 10-K, Nvidia said it has invested over $58.2 billion in research and development since its inception, citing the invention of the GPU in 1999, the introduction of the CUDA programming model in 2006, and the acquisition of networking company Mellanox in 2020 as pivotal moments. The company employs a fabless manufacturing model, contracting wafer production to foundries including Taiwan Semiconductor Manufacturing Company and Samsung Electronics, and final assembly to partners including Hon Hai Precision Industry and Wistron Corporation. (Nvidia Corporation, Form 10-K filed with the U.S. Securities and Exchange Commission, 26 February 2025)

How It Makes Money: Two Segments

In its 10-K for fiscal year 2025, Nvidia reported its results in two segments. The Compute & Networking segment encompasses data centre accelerated computing platforms, AI solutions and software, networking, automotive platforms for autonomous and electric vehicles, the Jetson embedded platform for robotics, and the DGX Cloud computing service. The Graphics segment covers GeForce GPUs for gaming and PCs, the GeForce NOW game-streaming service, Quadro/NVIDIA RTX GPUs for enterprise workstations, virtual GPU software for cloud-based visual computing, automotive infotainment platforms and Omniverse Enterprise software for industrial digital twins. (Nvidia Corporation, Form 10-K filed with the U.S. Securities and Exchange Commission, 26 February 2025)

The company also reports revenue by four end markets: Data Centre, Gaming, Professional Visualisation, and Automotive. In its earnings press release for the fourth quarter and full fiscal year 2025 (fiscal year ended 26 January 2025), Nvidia stated that Data Centre full-year revenue rose 142% to a record $115.2 billion. Gaming full-year revenue rose 9% to $11.4 billion. Professional Visualisation full-year revenue rose 21% to $1.9 billion. Automotive full-year revenue rose 55% to $1.7 billion. All figures are stated in US dollars. (Nvidia Corporation, Q4 fiscal 2025 earnings press release (Form 8-K), U.S. Securities and Exchange Commission, 26 February 2025)

Within the Data Centre market, Nvidia's compute offerings span GPUs, CPUs and data processing units, while its networking offerings include InfiniBand and Ethernet platforms, network adapters, cables, switch chips and software. The company also derives revenue from paid software licences, including NVIDIA AI Enterprise โ€” described in the 10-K as "a comprehensive suite of enterprise-grade AI software" โ€” and from DGX Cloud, a managed AI-training-as-a-service platform. The 10-K noted that NVIDIA AI Enterprise includes NIM microservices, the NeMo model development suite and AI Blueprints for enterprise AI agents. (Nvidia Corporation, Form 10-K filed with the U.S. Securities and Exchange Commission, 26 February 2025)

Reported Financial Results

In its earnings press release dated 26 February 2025, Nvidia reported that total revenue for fiscal year 2025 (the year ended 26 January 2025) was $130,497 million, up 114% from $60,922 million in the prior fiscal year. GAAP gross margin was 75.0%, up 2.3 percentage points year on year. GAAP operating income was $81,453 million, up 147%, and GAAP net income was $72,880 million, up 145%. GAAP diluted earnings per share was $0.89 for the fourth quarter and $2.94 for the full fiscal year โ€” with per-share figures adjusted to reflect the ten-for-one stock split effective 7 June 2024. (Nvidia Corporation, Q4 fiscal 2025 earnings press release (Form 8-K), U.S. Securities and Exchange Commission, 26 February 2025)

On 19 November 2025, Nvidia reported financial results for the third quarter of fiscal 2026, the quarter ended 26 October 2025. Revenue was $57,006 million, up 22% from the prior quarter and up 62% from the same quarter of the prior fiscal year. Data Centre revenue in the quarter was a record $51.2 billion, up 25% from the prior quarter and up 66% year on year. Gaming revenue was $4.3 billion, down 1% from the prior quarter and up 30% year on year. GAAP net income for the quarter was $31,910 million, with GAAP diluted earnings per share of $1.30. For the nine months ended 26 October 2025, cumulative revenue was $147,811 million. (Nvidia Corporation, Q3 fiscal 2026 earnings press release (Form 8-K), U.S. Securities and Exchange Commission, 19 November 2025)

In the same Q3 fiscal 2026 press release, Nvidia said that during the first nine months of fiscal 2026 it returned $37.0 billion to shareholders through share repurchases and cash dividends, and that as of the end of the third quarter it had $62.2 billion remaining under its share repurchase authorisation. The company stated that its outlook for the fourth quarter of fiscal 2026 is revenue of $65.0 billion, plus or minus 2%, and GAAP gross margin of 74.8%, plus or minus 50 basis points. These are the company's own published expectations and not a third-party forecast. (Nvidia Corporation, Q3 fiscal 2026 earnings press release (Form 8-K), U.S. Securities and Exchange Commission, 19 November 2025)

Stated Strategy

In its fiscal 2025 Form 10-K, Nvidia set out four strategic priorities. First, advancing its accelerated computing platform by pursuing performance gains across silicon, systems, networking, software and algorithms โ€” described as a "full-stack innovation approach." Second, extending its technology and platform leadership in AI, covering both training and inferencing, through its full-stack data-centre-scale compute and networking solutions, and through software including DGX Cloud and NVIDIA AI Enterprise. Third, extending its technology and platform leadership in computer graphics, integrating AI with graphics via tools such as DLSS and the RTX platform. Fourth, advancing its autonomous vehicle platform under the DRIVE brand through partnerships with automakers, tier-one suppliers and technology firms. (Nvidia Corporation, Form 10-K filed with the U.S. Securities and Exchange Commission, 26 February 2025)

The 10-K described a fifth element: leveraging intellectual property through licence and development agreements that allow customers and partners to incorporate Nvidia's technology directly into their own products. The company also highlighted its Inception programme for startups, partnerships with hundreds of universities and the Deep Learning Institute as mechanisms for expanding its developer ecosystem, which it stated comprised over 5.9 million CUDA developers worldwide. (Nvidia Corporation, Form 10-K filed with the U.S. Securities and Exchange Commission, 26 February 2025)

In the Q3 fiscal 2026 earnings press release, the company noted the launch of the Blackwell GPU architecture across data centre, gaming and professional visualisation products. It said the architecture had "achieved the highest performance and best overall efficiency in the SemiAnalysis InferenceMAX benchmarks" according to that benchmark's publisher. The press release also noted an announced investment of ยฃ2 billion in the United Kingdom market and a partnership with Deutsche Telekom to launch what the two companies described as "the world's first Industrial AI Cloud" for German industrial customers. (Nvidia Corporation, Q3 fiscal 2026 earnings press release (Form 8-K), U.S. Securities and Exchange Commission, 19 November 2025)

Where It Operates

Nvidia is incorporated in Delaware and headquartered in Santa Clara, California. Its 10-K stated that its supply chain is "mainly concentrated in the Asia-Pacific region," with wafer production at TSMC's and Samsung's foundries, memory purchased from SK Hynix, Micron Technology and Samsung, and assembly and testing carried out by contract manufacturers. It sells through a global network of original equipment manufacturers, original device manufacturers, system integrators, distributors and cloud service providers. (Nvidia Corporation, Form 10-K filed with the U.S. Securities and Exchange Commission, 26 February 2025)

The Q3 fiscal 2026 press release detailed activity across multiple geographies, including partnerships with the South Korean government and industrial groups including Hyundai Motor Group, Samsung Electronics and SK Group to expand AI infrastructure with over a quarter-million NVIDIA GPUs; the first NVIDIA Blackwell wafer produced at TSMC's Arizona facility in the United States; and work with partners to build what the press release called the United Kingdom's "next generation of AI infrastructure." In fiscal year 2025, the company stated it opened its first research and development centre in Vietnam. Its shares are listed on the Nasdaq Global Select Market under the ticker symbol NVDA. (Nvidia Corporation, Q3 fiscal 2026 earnings press release (Form 8-K), U.S. Securities and Exchange Commission, 19 November 2025; Nvidia Corporation, Q4 fiscal 2025 earnings press release (Form 8-K), U.S. Securities and Exchange Commission, 26 February 2025)

The Friday Digest

Enjoyed this? Get the week's top business stories, every Friday.

Free. One email a week. Unsubscribe in one click.

TPC Editorial Desk

Written by

TPC Editorial Desk

The Platinum Capital editorial desk

Articles on The Platinum Capital are produced by the TPC editorial desk using AI tools and are fact-checked against public sources. Profiles are based on information provided by the subject. Reach out at info@theplatinumcapital.com.