What changed in this filing

/stocks — the workspaceLOADING

Item 1. Business

2026-02-25 compared with 2025-02-26 · 127 added, 188 removed, 189 unchanged (63% of the section changed)

NVIDIA pioneered accelerated computing to help solve the most challenging computational problems.

NVIDIA is now a full-stack computing infrastructure company with data-center-scale offerings that are reshaping industry.

Our full-stack includes the foundational CUDA programming model that runs on all NVIDIA GPUs, as well as hundreds of domain-specific software libraries, software development kits, or SDKs, and Application Programming Interfaces, or APIs.

This deep and broad software stack accelerates the performance and eases the deployment of NVIDIA accelerated computing for computationally intensive workloads such as artificial intelligence, or AI, model training and inference, data analytics, scientific computing, and 3D graphics, with vertical-specific optimizations to address industries ranging from healthcare and telecom to automotive and manufacturing.

Our data-center-scale offerings are comprised of compute and networking solutions that can scale to tens of thousands of GPU-accelerated servers interconnected to function as a single giant computer;

this type of data center architecture and scale is needed for the development and deployment of modern AI applications.

NVIDIA is now a data center scale AI infrastructure company reshaping all industries.

Our technology stack includes the foundational NVIDIA CUDA development platform that runs on all NVIDIA GPUs, as well as hundreds of domain-specific software libraries, frameworks, algorithms, software development kits, or SDKs, and application programming interfaces, or APIs.

This deep and broad software stack accelerates the performance and facilitates the deployment of NVIDIA accelerated computing for computationally intensive workloads such as artificial intelligence, or AI, model training and inference, data analytics, scientific computing, robotics, and 3D graphics, with vertical-specific optimizations to address industries ranging from healthcare and telecom to automotive and manufacturing.

Introduced with the Blackwell architecture, our data-center-scale offerings feature extreme co-design where the infrastructure’s chips, networking, systems, software, and algorithms are holistically architected and optimized to maximize performance and scale.

Hundreds of thousands of GPUs can be interconnected to function as a single giant computer.

This type of data center architecture and scale is needed for the development and deployment of modern AI and accelerated computing applications.

The GPU was initially used to simulate human imagination, enabling the virtual worlds of video games and films.

Today, it also simulates human intelligence, enabling a deeper understanding of the physical world.

Its parallel processing capabilities, supported by thousands of computing cores, are essential for deep learning algorithms.

This form of AI, in which software writes itself by learning from large amounts of data, can serve as the brain of computers, robots, and self-driving cars that can perceive and understand the world.

Today, it also simulates human intelligence, enabling a deeper understanding of language, science, and the physical world.

Its parallel processing capabilities, supported by tens of thousands of computing cores, are essential for deep learning algorithms.

This form of AI, in which software writes itself by learning from large amounts of data, can serve as the brain of computers, robots, and self-driving cars that can perceive, understand and reason about the world.

GPU-powered AI solutions are being developed by thousands of enterprises to deliver services and products that would have been immensely difficult or even impossible with traditional coding.

Examples include generative AI, which can create new content such as text, code, images, audio, video, molecule structures, and recommendation systems, which can recommend highly relevant content such as products, services, media, or ads using deep neural networks trained on vast datasets that capture the user's preferences.

NVIDIA has a platform strategy, bringing together hardware, systems, software, algorithms, libraries, and services to create unique value for the markets we serve.

While the computing requirements of these end markets are diverse, we address them with a unified underlying architecture leveraging our GPUs and networking and software stacks.

The programmable nature of our architecture allows us to support several multi-billion-dollar end markets with the same underlying technology by using a variety of software stacks developed either internally or by third-party developers and partners.

The large and growing number of developers and installed base across our platforms strengthens our ecosystem and increases the value of our platform to our customers.

Examples include generative AI, which can create new content such as text, code, images, audio, video, molecule structures, and recommendation systems;

and agentic AI where systems of AI models work in concert to automatically complete a task.

NVIDIA has a platform strategy, bringing together hardware, systems, software, algorithms, libraries, AI models and training data sets, and services to create unique value for the markets we serve.

While the computing requirements of these end markets are diverse, we address them with a unified underlying programmable architecture allowing us to support several multi-billion-dollar end markets with the same underlying technology by using a variety of software stacks developed either internally or by third-party developers and partners.

The large and growing number of developers and installed base across our platforms strengthens our ecosystem and increases the value of our platform for our customers.

Innovation is at our core.

1 unchanged sentence

Our invention of the GPU in 1999 sparked the growth of the PC gaming market and redefined computer graphics.

With our introduction of the CUDA programming model in 2006, we opened the parallel processing capabilities of our GPU to a broad range of compute-intensive applications, paving the way for the emergence of modern AI.

With our introduction of CUDA in 2006, we opened the parallel processing capabilities of our GPU to a broad range of compute-intensive applications, paving the way for the emergence of modern AI.

In 2012, the AlexNet neural network, trained on NVIDIA GPUs, won the ImageNet computer image recognition competition, marking the “Big Bang” moment of AI.

We introduced our first Tensor Core GPU in 2017, built from the ground-up for the new era of AI, and our first autonomous driving system-on-chips, or SoC, in 2018.

Our acquisition of Mellanox in 2020 expanded our innovation canvas to include networking, enabled our platforms to be data center scale, and led to the introduction of a new processor class – the data processing unit, or DPU.

Over the past 5 years, we have built full software stacks that run on top of our GPUs and CUDA to bring AI to the world’s largest industries, including NVIDIA DRIVE stack for autonomous driving, Clara for healthcare, and Omniverse for industrial digitalization;

and introduced the NVIDIA AI Enterprise software – essentially an operating system for enterprise AI applications.

Our acquisition of Mellanox in 2020 expanded our offerings to include networking, enabled our platforms to be data center scale, and led to the introduction of a new processor class – the data processing unit, or DPU.

In 2023, we introduced our first data center CPU, Grace, built for giant-scale AI and high-performance computing, or HPC.

In 2024, we launched the NVIDIA Blackwell architecture – connecting 36 Grace CPUs and 72 Blackwell GPUs in a data center scale, liquid-cooled design – for real-time trillion-parameter inference and training.

In fiscal year 2026, we launched and scaled the NVIDIA Blackwell Ultra platform, optimized for agentic, reasoning, and physical AI.

Building on the architectural breakthroughs of Blackwell and leveraging Dynamo inference software, it delivers a significant increase in token throughput and reduction in cost per token compared to the Hopper generation.

More recently, in support of market development, we have accelerated the release cadence of our open AI model platforms including NVIDIA Nemotron for agentic AI and Cosmos for physical AI.

With a strong engineering culture, we drive fast, yet harmonized, product and technology innovations in all dimensions of computing including silicon, systems, networking, software and algorithms.

More than half of our engineers work on software.

The world’s leading cloud service providers, or CSPs, and consumer internet companies use our data center-scale accelerated computing platforms to enable, accelerate, develop, or enrich the services and offerings they deliver to billions of end users, including AI solutions and assistants, AI foundation models, search, recommendations, social networking, online shopping, live video, and translation.

All major cloud service providers, or CSPs, AI model makers, and enterprises use our data center-scale infrastructure and computing platforms to accelerate the services and offerings they deliver to billions of end users and customers, including AI solutions and assistants, AI foundation models, advertising, search, recommendation systems, social

networking, data processing, online shopping, live video, and translation.

AI model makers use our infrastructure and software hosted at CSPs to develop, build and run AI models, product offerings, and services.

Enterprises and startups across a broad range of industries use our accelerated computing platforms to build new generative and agentic AI-enabled products and services, and/or to dramatically accelerate and reduce the costs of their workloads and workflows.

4 unchanged sentences

Researchers and developers use our computing solutions to accelerate a wide range of important applications, from simulating molecular dynamics to climate forecasting.

With support for more than 4,400 applications, NVIDIA computing enables some of the most promising areas of discovery, from climate prediction to materials science and from wind tunnel simulation to genomics.

With support for 6,000 applications, NVIDIA computing enables some of the most promising areas of discovery, from climate prediction to materials science and from wind tunnel simulation to genomics.

Including GPUs and networking, NVIDIA powers over 78% of the supercomputers on the global TOP500 list, including 9 of the top 10 systems on the Green500 list.

Gamers choose NVIDIA GPUs to enjoy immersive, increasingly cinematic virtual worlds.

In addition to serving the growing number of gamers, the market for PC GPUs is expanding because of the burgeoning population of live streamers, broadcasters, artists, and creators.

With the advent of generative AI, we expect a broader set of PC users to choose NVIDIA GPUs for running generative AI applications locally on their PC, which is critical for privacy, latency, and cost-sensitive AI applications.

Professional artists, architects and designers use NVIDIA partner products accelerated with our GPUs and software platform for a range of creative and design use cases, such as creating visual effects in movies or designing buildings and products.

In addition, generative AI is expanding the market for our workstation-class GPUs, as more enterprise customers develop and deploy AI applications with their data on-premises.

In addition to serving the growing number of gamers, the market for PC GPUs is expanding because of the growing population of live streamers, broadcasters, artists, and creators.

With the advent of generative and agentic AI, we expect a broader set of PC users to choose NVIDIA GPUs for running these applications locally on their PC, which is critical for privacy, latency, and cost-sensitive AI applications.

Professional artists, architects and designers use NVIDIA partner products accelerated with our GPUs and software platform for a range of creative, engineering, and design use cases, such as creating visual effects in movies or designing buildings and products.

In addition, generative and agentic AI is expanding the market for our workstation-class GPUs, as more enterprise customers develop and deploy AI applications with their data on-premises.

Headquartered in Santa Clara, California, NVIDIA was incorporated in California in April 1993 and reincorporated in Delaware in April 1998.

1 unchanged sentence

We report our business results in two segments.

The Compute & Networking segment includes our Data Center accelerated computing platforms and AI solutions and software;

automotive platforms and autonomous and electric vehicle solutions;

Jetson for robotics and other embedded platforms;

and DGX Cloud computing services.

The Graphics segment includes GeForce GPUs for gaming and PCs, the GeForce NOW game streaming service and related infrastructure, and solutions for gaming platforms;

Quadro/NVIDIA RTX GPUs for enterprise workstation graphics;

virtual GPU, or vGPU, software for cloud-based visual and virtual computing;

automotive platforms for infotainment systems;

and Omniverse Enterprise software for building and operating industrial AI and digital twin applications.

We specialize in markets where our computing platforms can provide tremendous acceleration for applications.

The Compute & Networking segment includes our Data Center accelerated computing and networking platforms and AI solutions and software, and Automotive platforms and autonomous and electric vehicle solutions including software.

The Graphics segment includes GeForce GPUs for gaming and PCs, and Quadro/NVIDIA RTX GPUs for enterprise workstation graphics.

We specialize in markets where our computing and AI infrastructure platforms can provide tremendous acceleration for applications.

These platforms incorporate processors, interconnects, software, algorithms, systems, and services to deliver unique value.

1 unchanged sentence

Data Center, Gaming, Professional Visualization, and Automotive.

The NVIDIA Data Center platform is focused on accelerating the most compute-intensive workloads, such as AI, data analytics, graphics, and scientific computing, delivering significantly better performance and power efficiency relative to conventional CPU-only approaches.

The NVIDIA Data Center platform is focused on accelerating compute-intensive workloads, such as AI, data processing, graphics, robotics, and scientific computing, delivering superior total cost of ownership relative to conventional CPU-only approaches.

It is deployed in cloud, hyperscale, on-premises and edge data centers.

The platform consists of compute and networking offerings typically delivered to customers as systems, subsystems, or modules, along with software and services.

Our compute offerings include supercomputing platforms and servers, bringing together our energy efficient GPUs, CPUs, interconnects, and fully optimized AI and HPC software stacks.

In addition, they include NVIDIA AI Enterprise software;

our DGX Cloud service;

and a growing body of acceleration libraries, APIs, SDKs, and domain-specific application frameworks.

Our networking offerings include end-to-end platforms for InfiniBand and Ethernet, consisting of network adapters, cables, DPUs, switch chips and systems, as well as a full software stack.

This has enabled us to architect data center-scale computing platforms that can interconnect thousands of compute nodes with high-performance networking.

While historically the server was the unit of computing, as AI and HPC workloads have become extremely large spanning thousands of compute nodes, the data center has become the new unit of computing, with networking as an integral part.

Our customers include the world’s leading public cloud and consumer internet companies, thousands of enterprises and startups, and public sector entities.

The platform consists of data center compute and networking infrastructure offerings typically delivered to customers as rack-scale systems, subsystems, or modules, along with software and services.

Our Data Center infrastructure systems include supercomputing platforms and servers, bringing together our higher performance, energy efficient GPUs, CPUs, interconnects, and fully optimized AI and HPC software stacks.

In addition, they include a growing body of acceleration libraries, AI models and training data sets, APIs, SDKs, and domain-specific application frameworks.

Our networking offerings include NVLink interconnects and switches, end-to-end platforms for InfiniBand and Ethernet, consisting of network adapters, cables, DPUs, switch chips and systems, as well as software.

This has enabled us to architect data center-scale computing platforms that can interconnect up to hundreds of thousands of compute nodes with high-performance networking.

Fueled by an expansion in AI and HPC workloads, the data center has become the new unit of computing, with networking as an integral part.

In fiscal year 2026, we introduced NVIDIA NVLink Fusion to enable hyperscalers and custom ASIC designers to integrate custom CPUs and XPUs with our platform.

Our customers include all major public and private cloud providers, AI model makers, enterprises and startups, and public sector entities.

We work with industry leaders to help build or transform their applications and data center infrastructure.

Our direct customers include original equipment manufacturers, or OEMs, original device manufacturers, or ODMs, system integrators and distributors which we partner with to help bring our products to market.

Some of our direct customers include original equipment manufacturers, or OEMs, original device manufacturers, or

ODMs, system integrators and distributors which we partner with to help bring our products to market.

We also have partnerships in automotive, healthcare, financial services, manufacturing, retail, and technology among others, to accelerate the adoption of AI.

At the foundation of the NVIDIA accelerated computing platform are our GPUs, which excel at parallel workloads such as the training and inferencing of neural networks.

They are available in the NVIDIA accelerated computing platform and in

industry standard servers from every major cloud provider and server maker.

Beyond GPUs, our data center platform expanded to include DPUs in fiscal year 2022 and CPUs in fiscal year 2024.

We can optimize across the entire computing, networking and storage stack to deliver data center-scale computing solutions.

While our approach starts with powerful chips, what makes it a full-stack computing platform is our large body of software, including the CUDA parallel programming model, the CUDA-X collection of acceleration libraries, APIs, SDKs, and domain-specific application frameworks.

In addition to software delivered to customers as an integral part of our data center computing platform, we offer paid licenses to NVIDIA AI Enterprise, a comprehensive suite of enterprise-grade AI software and NVIDIA vGPU software for graphics-rich virtual desktops and workstations.

We also offer the NVIDIA DGX Cloud, a fully managed AI-training-as-a-service platform which includes cloud-based infrastructure and software for AI, customizable pretrained AI models, and access to NVIDIA experts.

These Data Center systems are extreme co-designed with the GPU, CPU, NVLink switch, DPU, NIC, and scale-out networking along with software stacks and algorithms to deliver data center-scale computing solutions.

While our approach starts with powerful chips, what makes it a full-stack computing platform is our large body of software, including the CUDA development platform, the CUDA-X collection of acceleration libraries, AI models and training data sets, APIs, SDKs, and domain-specific application frameworks.

In addition to software delivered to customers as an integral part of our data center computing and networking platform, we offer paid licenses to NVIDIA AI Enterprise, a comprehensive suite of enterprise-grade AI software and NVIDIA vGPU software for graphics-rich virtual desktops and workstations.

In fiscal year 2025, we launched the NVIDIA Blackwell architecture, a full set of data center scale infrastructure that includes GPUs, CPUs, DPUs, interconnects, switch chips and systems, and networking adapters.

Blackwell excels at processing cutting edge generative AI and accelerated computing workloads with market leading performance and efficiency.

Offered in a number of configurations, it can address the needs of customers across industries and a diverse set of AI and accelerated computing use cases.

Offered in a number of configurations, for customers across industries and a diverse set of AI and accelerated computing use cases.

In fiscal year 2026, we unveiled the NVIDIA Rubin platform, which is expected to commence production shipments in the second half of fiscal year 2027.

Built for agentic AI and reasoning, it excels at processing multi-step problem-solving and massive long-context workflows, delivering up to a 10x reduction in cost per token compared to Blackwell.

For physical AI, we provide an end-to-end platform spanning data center infrastructure, open models, systems, embedded compute modules, and software stacks to train, simulate, and deploy advanced automation and robotics solutions.

Gaming is the largest entertainment industry, with PC gaming as the predominant platform.

Many factors propel its growth, including new high production value games and franchises, the continued rise of competitive gaming, or eSports, social connectivity and the increasing popularity of game streamers, modders, or gamers who remaster games, and creators.

Many factors propel its growth, including new high production value games, the continued rise of eSports, social connectivity and the increasing popularity of game streamers, modders, or gamers who remaster games, and creators.

Our products for the gaming market include GeForce RTX GPUs for gaming desktop and laptop PCs, GeForce NOW cloud gaming service, as well as SoCs and development services for game consoles.

Our gaming platforms leverage our GPUs and sophisticated software to enhance the gaming experience with smoother, higher quality graphics.

We developed NVIDIA RTX to bring next generation graphics and AI to games.

NVIDIA RTX features ray tracing technology for real-time, cinematic-quality rendering.

Ray tracing, which has long been used for special effects in the movie industry, is a computationally intensive technique that simulates the physical behavior of light to achieve greater realism in computer-generated scenes.

NVIDIA RTX also features deep learning super sampling, or NVIDIA DLSS, our AI technology that boosts frame rates while generating beautiful, sharp images for games.

RTX GPUs will also accelerate a new generation of AI applications.

With an installed base of over 100 million AI capable PCs, more than 700 RTX AI-enabled applications and games, and a robust suite of development tools, RTX is already the AI PC leader.

Our products for the gaming market include GeForce RTX and GeForce GTX GPUs for gaming desktop and laptop PCs, GeForce NOW cloud gaming for playing PC games on underpowered devices, as well as SoCs and development services for game consoles.

In fiscal year 2025, we launched the NVIDIA Blackwell GeForce RTX 50 Series family of desktop and laptop GPUs.

The Blackwell architecture introduced neural graphics which combines AI models with traditional rendering to unlock a new era of graphics innovation.

The RTX 50 Series also features the next generation of our DLSS technology powered for the first time by a transformer model architecture.

Together these technologies help deliver up to a 2x leap in performance and stunning visual realism for PC gamers, developers, and creatives.

NVIDIA RTX features ray tracing technology for real-time, cinematic-quality rendering, and deep learning super sampling, or NVIDIA DLSS, our AI technology that boosts frame rates while generating high-quality images for games.

RTX GPUs also feature NVIDIA tensor core technology making them well suited to accelerate a new generation of on-device AI applications.

In fiscal year 2025, we announced the NVIDIA Blackwell GeForce RTX 50 Series family of desktop and laptop GPUs.

The Blackwell architecture introduced neural graphics which combines AI models with traditional rendering to boost game performance, image quality, and interactivity, as well as the next generation of our DLSS technology powered by a new transformer model architecture.

In fiscal year 2026, we launched and scaled Blackwell architecture for gaming and GeForce NOW.

Professional Visualization

We serve the Professional Visualization market by working closely with independent software vendors, or ISVs, to optimize their offerings for NVIDIA GPUs.

Our GPU computing platform enhances productivity and introduces new capabilities for critical workflows in many fields, such as design and manufacturing and digital content creation.

Design and manufacturing encompass computer-aided design, architectural design, consumer-products manufacturing, medical instrumentation, and aerospace.

Digital content creation includes professional video editing and post-production, special effects for films, and broadcast-television graphics.

Additionally, the infusion of generative AI into an increasing number of applications is giving rise to the need for the enhanced AI processing capabilities of our RTX GPUs.

The NVIDIA RTX platform makes it possible to render film-quality, photorealistic objects and environments with physically accurate shadows, reflections and refractions using ray tracing in real-time.

Many leading 3D design and content creation applications developed by our ecosystem partners now support RTX, allowing professionals to accelerate and transform their workflows with NVIDIA RTX GPUs and software.

We offer NVIDIA Omniverse as a development platform and operating system for building and running virtual world simulation applications, available as a software subscription for enterprise use and free for individual use.

Industrial enterprises are adopting Omniverse’s 3D and simulation technologies to digitalize their complex physical assets, processes, and environments – building digital twins of factories, real time 3D product configurators, testing and validating autonomous robots and vehicles, powered by NVIDIA accelerated computing infrastructure on-premises and in the cloud.

Our GPU computing platform enhances productivity and introduces new capabilities for critical workflows in many fields, such as design, engineering, and digital content creation across a wide range of industry verticals.

Additionally, the increasing number of generative and agentic AI applications is giving rise to the need for the enhanced AI and data processing capabilities of our RTX PRO GPUs.

Many leading 3D design and content creation applications developed by our ecosystem partners support RTX, allowing professionals to accelerate and transform their workflows with NVIDIA RTX PRO GPUs and software.

As these applications increasingly integrate AI, these GPUs are used and leverage the same Tensor Core technology found in our Data Center solutions.

Automotive is comprised of platform solutions for automated driving from the cloud to the car.

Leveraging our technology leadership in AI and building on our long-standing automotive relationships, we are delivering a complete end-to-end solution for the AV market under the DRIVE Hyperion brand.

We have demonstrated multiple applications of AI within the car:

AI can drive the car itself as a pilot in fully autonomous mode or it can also be a co-pilot, assisting the human driver while creating a safer driving experience.

We are working with several hundred partners in the automotive ecosystem including automakers, truck makers, tier-one suppliers, sensor manufacturers, automotive research institutions, HD mapping companies, and startups to develop and deploy AI systems for self-driving vehicles.

Our unified AI computing architecture starts with training deep neural networks using our Data Center computing solutions, and then running a full perception, fusion, planning, and control stack within the vehicle on the NVIDIA DRIVE Hyperion platform.

DRIVE Hyperion consists of the high-performance, energy efficient DRIVE AGX computing hardware running an in-vehicle operating system (DRIVE OS), a reference sensor set that supports full self-driving capability as well as an open, modular DRIVE software platform for autonomous driving, mapping, and parking services, and intelligent in-vehicle experiences.

In addition, we offer a scalable data center-based simulation solution based on NVIDIA Omniverse software to develop synthetic data for AI model training, as well as for testing and validating a self-driving platform.

Our unique end-to-end, software-defined approach is designed for continuous innovation and continuous development, enabling cars to receive over-the-air updates to add new features and capabilities throughout the life of a vehicle.

Leveraging our technology leadership in AI and building on long-standing relationships across several hundred automotive ecosystem partners, we are delivering a full stack end-to-end solution for the AV market under the DRIVE Hyperion platform.

This platform consists of development infrastructure, high-performance, energy efficient DRIVE AGX computing hardware

running an in-vehicle operating system (DRIVE OS), a reference sensor set that supports full self-driving capability as well as an open, modular DRIVE software platform for autonomous driving, mapping, and parking services, and intelligent in-vehicle experiences.

Business Strategies

2 unchanged sentences

Our accelerated computing platform can solve complex problems in significantly less time and with lower power consumption than alternative computational approaches.

Indeed, it can help solve problems that were previously deemed unsolvable.

It can help solve problems that were previously deemed unsolvable.

We work to deliver continued performance leaps that outpace Moore’s Law by leveraging innovation across the architecture, chip design, system, interconnect, algorithm, and software layers.

9 unchanged sentences

GPUs are uniquely suited to AI, and we will continue to add AI-specific features to our GPU architecture to further extend our leadership position.

In addition, we offer DGX Cloud, a fully managed AI-training-as-a-service platform, along with NVIDIA AI Enterprise—a comprehensive software suite designed to simplify the development and deployment of production-grade, end-to-end generative AI applications.

In addition, we offer NVIDIA AI Enterprise—a comprehensive software suite designed to simplify the development and deployment of production-grade, end-to-end generative AI applications.

NVIDIA AI Enterprise includes:

NVIDIA NIM, which delivers a 2.5x increase in token throughput using industry-leading open and proprietary models;

NVIDIA NeMo, a complete solution for curating, fine-tuning, evaluating, and safeguarding domain-adapted models;

NVIDIA NIM, which increases token throughput using industry-leading open and proprietary models;

NVIDIA NeMo, a complete solution for curating, fine-tuning, reinforcement learning, evaluating, and safeguarding domain-adapted models;

and AI Blueprints, pre-built, runnable templates that help enterprises build, optimize, and deploy AI agents while preserving privacy.

These tools enable organizations to securely develop and run AI applications on NVIDIA-accelerated infrastructure anywhere.

Our AI technology leadership is reinforced by our large and expanding ecosystem in a virtuous cycle.

Our AI technology leadership is reinforced by our large and expanding ecosystem.

Our computing platforms are available from virtually every major server maker and CSP, as well as on our own AI supercomputers.

There are over 7.5 million developers worldwide using CUDA and our other software tools to help deploy our technology in our target markets.

We evangelize AI through partnerships with hundreds of universities and thousands of startups through our Inception program.

We are the leader in accelerating and releasing open AI models which enterprises, sovereigns, and startups can leverage to develop and run applications on our platform.

We evangelize AI through partnerships with hundreds of universities and tens of thousands of startups through our Inception program.

Additionally, our Deep Learning Institute provides instruction on the latest techniques on how to design, train, and deploy neural networks in applications using our accelerated computing platform.

4 unchanged sentences

Our computer graphics platforms leverage AI end-to-end, from the developer tools and cloud services to the Tensor Cores included in all RTX-class GPUs.

For example, NVIDIA Avatar Cloud Engine, or ACE, is

a suite of technologies that help developers bring digital avatars to life with generative AI, running in the cloud or locally on the PC.

GeForce Experience enhances each gamer’s experience by optimizing their PC’s settings, as well as enabling the recording and sharing of gameplay.

Our Studio drivers enhance and accelerate a number of popular creative applications.

Blackwell GPUs’ advanced AI and neural rendering capabilities combined with NVIDIA’s world-class AI software stacks significantly accelerate AI workloads running locally on PCs.

Omniverse is real-time 3D design collaboration and virtual world simulation software that empowers artists, designers, and creators to connect and collaborate in leading design applications.

We also enable interactive graphics applications - such as games, movie and photo editing and design software - to be accessed by almost any device, almost anywhere, through our cloud platforms such as vGPU for enterprise and GeForce NOW for gaming.

Advancing the leading autonomous vehicle platform.

We believe the advent of autonomous vehicles, or AV, and electric vehicles, or EV, is revolutionizing the transportation industry.

The algorithms required for autonomous driving - such as perception, localization, and planning - are too complex for legacy hand-coded approaches and will use multiple neural networks instead.

Therefore, we provide an AI-based hardware and software solution, designed and implemented from the ground up based on automotive safety standards, for the AV and EV market under the DRIVE brand, which we are bringing to market through our partnerships with automotive OEMs, tier-1 suppliers, and start-ups.

The algorithms required for autonomous driving - such as reasoning, perception, localization, and planning - are too complex for legacy hand-coded approaches and will use multiple neural networks instead.

Therefore, we provide an AI-based hardware and software solution, designed and implemented from the ground up based on automotive safety standards, for the AV and EV market under the DRIVE brand, which we are bringing to market through our partnerships across the transportation industry including with automotive OEMs, mobility service providers, robotaxis, tier-1 suppliers, and start-ups.

Our AV solution also includes the GPU-based hardware required to train the neural networks before their in-vehicle deployment, as well as to re-simulate their operation prior to any over-the-air software updates.

9 unchanged sentences

We also employ a team of application engineers and solution architects to provide pre-sales assistance to our partner network in designing, testing, and qualifying system designs that incorporate our platforms.

For example, our solution architects work with CSPs to provide pre-sales assistance to enable our customers to optimize their hardware and software infrastructure for generative AI and LLM training and deployment.

They also work with foundation model and enterprise software developers to enable our customers to optimize the training and fine-tuning of their models and services, and with enterprise end-users, often in collaboration with their global system integrator of choice, to fine-tune models and build AI applications.

13 unchanged sentences

As a result, we can focus our resources on product design, quality assurance, marketing, and customer support.

of growth, we may place non-cancellable inventory orders for certain product components in advance of our historical lead times, pay premiums, or provide deposits to secure future supply and capacity and may need to continue to do so.

In periods of growth, we may place non-cancellable inventory orders for certain product components in advance of our historical lead times, pay premiums, or provide deposits to secure future supply and capacity and may need to continue to do so.

We have expanded our supplier relationships to build redundancy and resilience in our operations to provide long-term manufacturing capacity aligned with growing customer demand.

Our supply chain is mainly concentrated in the Asia-Pacific region.

While currently our supply chain is mainly concentrated in Asia, we are expanding into the U.S.

and Latin America.

We utilize foundries, such as Taiwan Semiconductor Manufacturing Company Limited, or TSMC, and Samsung Electronics Co., Ltd., or Samsung, to produce our semiconductor wafers.

4 unchanged sentences

We believe that the principal competitive factors in this market are performance, breadth of product offerings, access to customers and partners and distribution channels, software support, conformity to industry standard APIs, manufacturing capabilities, processor pricing, and total system costs.

We believe that our ability to remain competitive will depend on how well we are able to anticipate the features and functions that customers and partners will demand and whether we are able to deliver consistent volumes of our products at acceptable levels of quality and at competitive prices.

We believe that our ability to remain competitive will depend on how well we are able to anticipate the features and functions that customers and partners will

demand and whether we are able to deliver consistent volumes of our products at acceptable levels of quality and at competitive prices.

We expect competition to increase from both existing competitors and new market entrants with products that may be lower priced than ours or may provide better performance or additional features not provided by our products.

8 unchanged sentences

• suppliers of Arm-based CPUs and companies that incorporate hardware and software for CPUs as part of their internal solutions or platforms, such as Amazon, Huawei, and Microsoft;

• suppliers of hardware and software for SoC products that are used in servers or embedded into automobiles, autonomous machines, and gaming devices, such as Ambarella, Inc., AMD, Broadcom, Inc., or Broadcom, Intel, Qualcomm Incorporated, Renesas Electronics Corporation, and Samsung, or companies with internal teams designing SoC products for their own products and services, such as Tesla, Inc.;

• networking products consisting of switches, network adapters (including DPUs), and cable solutions (including optical modules) include such as AMD, Arista Networks, Broadcom, Cisco Systems, Inc., Hewlett Packard Enterprise Company, Huawei, Intel, Lumentum Holdings Inc., and Marvell Technology, Inc.

as well as internal teams of system vendors and large cloud services companies.

• suppliers of hardware and software for SoC products that are used in servers or embedded into automobiles, autonomous machines, and gaming devices, such as Ambarella, Inc., AMD, Broadcom, Intel, Qualcomm Incorporated, Renesas Electronics Corporation, and Samsung, or companies with internal teams designing SoC products for their own products and services, such as Tesla, Inc.;

Patents and Proprietary Rights

We rely primarily on a combination of patents, trademarks, trade secrets, employee and third-party nondisclosure agreements, and licensing arrangements to protect our IP in the United States and internationally.

Our currently issued patents have expiration dates from February 2025 to June 2045.

Our currently issued patents have expiration dates from March 2026 to June 2045.

We have numerous patents issued, allowed, and pending in the United States and in foreign jurisdictions.

13 unchanged sentences

In August 2022, the U.S.

government, or the USG, announced licensing requirements that, with certain exceptions, impact exports to China (including Hong Kong and Macau) and Russia of our A100 and H100 integrated circuits, DGX or any other systems or boards which incorporate A100 or H100 integrated circuits.

In July 2023, the USG informed us of an additional licensing requirement for a subset of A100 and H100 products destined to certain customers and other regions, including some countries in the Middle East.

In October 2023, the USG announced new and updated licensing requirements that became effective in our fourth quarter of fiscal year 2024 for exports to China and Country Groups D1, D4, and D5 (including but not limited to Saudi Arabia, the United Arab Emirates, and Vietnam, but excluding Israel) of our products exceeding certain performance thresholds, including, but not limited to, the A100, A800, H100, H800, L4, L40, L40S and RTX 4090.

government, or USG, announced export restrictions and export licensing requirements targeting China’s semiconductor and supercomputing industries.

These restrictions impacted exports of certain chips, as well as software, hardware, equipment and technology used to develop, produce and manufacture certain chips to China (including Hong Kong and Macau) and Russia, and specifically impact our A100 and H100 integrated circuits, DGX or any other systems or boards which incorporate A100 or H100 integrated circuits.

In July 2023, the USG also informed us of an additional licensing requirement for a subset of A100 and H100 products destined to certain customers and other regions, including some countries in the Middle East.

In October 2023, the USG announced new and updated licensing requirements for exports to China and Country Groups D:1, D:4, and D:5 (including but not limited to Saudi Arabia, the United Arab Emirates, and Vietnam, but excluding Israel) of our products exceeding certain performance thresholds, including, but not limited to, the A100, A800, H100, H800, L4, L40, L40S RTX 4090, GB200 NVL72, and B200.

The licensing requirements also apply to the export of products exceeding certain performance thresholds to a party headquartered in, or with an ultimate parent headquartered in, Country Group D5, including China.

On October 23, 2023, the USG informed us that the licensing requirements were effective immediately for shipments of our A100, A800, H100, H800, and L40S products (removing the grace period granted by the official rule).

Blackwell systems, such as GB200 NVL 72 and NVL 36 as well as B200 are also subject to these requirements and therefore require a license for any shipment to certain entities and to China and Country Groups D1, D4 and D5, excluding Israel.

To date, we have not received licenses to ship these restricted products to China.

Additionally, we understand that partners and customers have also not received a license to ship these restricted products.

On January 15, 2025, the USG published the “AI Diffusion” IFR in the Federal Register.

After a 120-day delayed compliance period, the IFR will, unless modified, impose a worldwide licensing requirement on all products classified under Export Control Classification Numbers, or ECCNs, 3A090.a, 4A090.a, or corresponding .z ECCNs, including all related software and technology.

Any system that incorporates one or more of the covered integrated circuits, or ICs, (including but not limited to NVIDIA DGX, HGX, and MGX systems) will be covered by the new licensing requirement.

The licensing requirement will include future NVIDIA ICs, boards, or systems classified with ECCN 3A090.a or 4A090.a, or corresponding .z ECCNs, achieving certain total processing performance and/or performance density.

Unless a license exception is available, the worldwide licensing requirements will apply to the following NVIDIA products, and any others we develop that meet the characteristics of 3A090.a or 4A090.a, including but not limited to:

A100, A800, H100, H200, H800, B100, B200, GB200, L4, L40S, and RTX 6000 Ada.

Our competitive position has been harmed by the existing export controls, and our competitive position and future results may be further harmed, over the long term, if there are further changes in the USG’s export controls.

Given the increasing strategic importance of AI and rising geopolitical tensions, the USG has changed and may again change the export control rules at any time and further subject a wider range of our products to export restrictions and licensing requirements, negatively impacting our business and financial results.

In the event of such change, we may be unable to sell our inventory of such products and may be unable to develop replacement products not subject to the licensing requirements, effectively excluding us from all or part of the China market, as well as other impacted markets, including the Middle East and countries designated “Tier 2” by the AI Diffusion IFR.

In addition to export controls, the USG may impose restrictions on the import and sale of products that incorporate technologies developed or manufactured in whole or in part in China.

For example, the USG is considering restrictions on the import and sale of certain automotive products in the United States, which if adopted and interpreted broadly, could impact our ability to develop and supply solutions for our automotive customers.

While we work to enhance the resiliency and redundancy of our supply chain, which is currently concentrated in the Asia-Pacific region, new and existing export controls or changes to existing export controls could limit alternative manufacturing locations and negatively impact our business.

In April 2025, the USG informed us that it requires a license for export to China (including Hong Kong and Macau) and D:5 countries, or to companies headquartered or with an ultimate parent therein, of our H20 integrated circuits and any other circuits achieving the H20’s memory bandwidth, interconnect bandwidth, or combination thereof.

As a result of these requirements, we incurred a $4.5 billion charge in the first quarter of fiscal year 2026 associated with H20 for excess inventory and purchase obligations, as the demand for H20 products diminished.

In August 2025, the USG granted licenses that would allow us to ship certain H20 products to certain China-based customers.

We generated approximately $60 million in H20 revenue under those licenses.

USG officials expressed an expectation that the USG will receive 15% or more of the revenue generated from licensed sales of our products, but the USG did not publish a regulation codifying such requirement.

In February 2026, the USG granted a license that would allow us to ship small amounts of H200 products to specific China-based customers.

To date, we have not generated any revenue under the H200 licensing program, and do not yet know whether any imports will be allowed into China.

The license requires that the H200s go through an inspection process in the United States prior to any shipment to the customer.

As a result, any H200 shipped under the new licensing program will be subject to a 25% tariff upon importation into the United States.

In the event that we are able to sell licensed products into the China market, we may not be able to pass along all or any of the tariff to our customers, and may be subject to litigation, increased costs, and a harmed competitive position.

The export controls applicable to China are complex and address a variety of parameters, including the total processing performance of a chip, the “performance density” of a chip, the interconnect bandwidth of a chip, and the memory bandwidth of a chip.

Under the current rules and geopolitical landscape, we are unable to create and deliver a competitive product for China’s data center market that receives approval from both the USG and the Chinese government.

As of the end of fiscal year 2026, we were effectively foreclosed from competing in China's data center computing/compute market, and our effective foreclosure from the China market helped our competitors build larger developer and customer ecosystems to challenge us worldwide.

Unless we are able to return with a product that meets the approval of both the USG and the Chinese government, our lost opportunity and the benefit to our competitors will have a material and adverse impact on our business, operating results, and financial condition.

In addition to controls targeting D:1, D:4 and D:5 countries, the USG has also imposed worldwide export controls impacting our products, and may impose additional controls in the future.

In January 2025, the USG published the AI Diffusion IFR in the Federal Register.

The IFR would have imposed a worldwide licensing requirement on our data center products, such as our H200, GB200 and GB300.

The AI Diffusion IFR would have divided the world into three tiers, relegating most countries to “Tier 2” status, and would have created a complex and burdensome scheme for licensing approvals.

In May 2025, the USG announced that it would rescind the AI Diffusion IFR and implement a replacement rule.

The scope, timing, and requirements of the forthcoming rule remain uncertain.

The replacement rule may impose new restrictions on our products or operations and/or add license requirements that could have a material impact on our business, operating results, and financial condition.

For example, in October 2025, the Senate passed the “GAIN AI Act” in the NDAA.

The GAIN AI Act would restrict the Trump Administration’s ability to adapt the Biden Administration’s export control rules, and could also allow private U.S.

persons to review and overturn licensing and foreign policy decisions made by the Trump Administration.

Our competitive position has been harmed by export controls, and our competitive position and future results will be further harmed, over the long term, if the restrictions remain in place or are expanded in geographic, customer, or product scope, if customers purchase product from competitors, if customers develop their own internal solution, if we are unable to provide contractual warranty or other extended service obligations, if the USG does not grant licenses in a timely manner or denies licenses to significant customers or if we incur significant transition costs.

The licensing process may not be resolved before significant business opportunities evaporate.

Even if the USG grants any requested licenses, the licenses have already and may in the future be temporary, impose burdensome conditions regarding the installation, maintenance, and use of such products, or include financial or economic requirements that we or our customers or end users cannot or choose not to fulfill.

The licensing requirements have already and may in the future benefit certain of our competitors, as the licensing process will make our pre-sale and post-sale technical support efforts more cumbersome and less certain and encourage customers in China, the Middle East, and other regions to pursue alternatives to our products, including semiconductor suppliers based in China, Europe, and Israel.

Additionally, restrictions imposed by the Chinese government on the duration of gaming activities and access to games may adversely affect our Gaming revenue, and even if we are able to participate in the China data center compute market, increased oversight of digital platform companies may adversely affect our Data Center revenue.

The Chinese government has encouraged customers to purchase from our China-based competitors and discouraged customers from purchasing, importing, or using our data center products, including any China-specific product designed to comply with U.S.

export controls.

While we work to enhance the resiliency and redundancy of our supply chain, which is currently concentrated in Asia, new and existing export controls or changes to existing export controls could limit alternative manufacturing locations and negatively impact our business.

Refer to “Item 1A.

4 unchanged sentences

Risk Factors” for a discussion of these potential impacts.

Sustainability and Governance

NVIDIA invents computing technologies that improve lives and address global challenges.

Our goal is to integrate sound environmental, social, and corporate governance principles and practices into every aspect of the Company.

The Nominating and Corporate Governance Committee of our Board of Directors is responsible for reviewing and discussing with management our practices related to sustainability and corporate governance.

We assess our programs annually in consideration of stakeholder expectations, market trends, and business risks and opportunities.

These issues are important for our continued business success and reflect the topics of highest concern to NVIDIA and our stakeholders.

The following section and the Human Capital Management Section below provide an overview of our principles and practices.

More information can be found on our website and in our annual Sustainability Report.

Information contained on our website or in our annual Sustainability Report is not incorporated by reference into this or any other report we file with the Securities and Exchange Commission, or the SEC.

Refer to “Item 1A.

Risk Factors” for a discussion of risks and uncertainties we face related to sustainability.

Climate Change

In the area of environmental sustainability, we work to address our climate impacts across our product lifecycle and assess related risks, including current and emerging regulations and market impacts.

In May 2024, we published metrics related to our environmental impact for fiscal year 2024.

Fiscal year 2025 metrics are expected to be published in the first half of fiscal year 2026.

There has been no material impact to our capital expenditures, results of operations or competitive position associated with global environmental sustainability regulations, compliance, or costs from sourcing renewable energy.

We committed to purchase or generate enough renewable energy to match 100% of our global electricity usage for offices and data centers under our operational control starting with our fiscal year 2025.

In fiscal year 2024, we made progress towards this goal and increased the percentage of our electricity use matched by renewable energy to 76%.

By the end of fiscal year 2026, we also aim to engage manufacturing suppliers comprising at least 67% of NVIDIA’s scope 3 category 1 GHG emissions with the goal of effecting supplier adoption of science-based targets.

Whether it is creation of technology to power next-generation laptops or designs to support high-performance supercomputers, improving energy efficiency is important in our research, development, and design processes.

GPU-accelerated computing is inherently more energy efficient than traditional computing for many workloads because it is optimized for throughput, performance per watt, and certain AI workloads.

We continue to have a strong presence on the Green500 list of the most energy-efficient systems—we powered 8 of the top 10 most energy efficient systems, including the top supercomputer, on the November 2024 Green500 list.

We launched our Earth-2 initiative to create a digital twin of the Earth on NVIDIA AI and NVIDIA Omniverse platforms.

Earth-2 will enable scientists, companies, and policy makers to do ultra-high-resolution predictions of the impact of climate change and explore mitigation and adaptation strategies.

Human Capital Management

We believe that our employees are our greatest assets, and they play a key role in creating long-term value for our stakeholders.

As of the end of fiscal year 2026, we had approximately 42,000 employees in 38 countries;

31,000 were engaged in research and development and 11,000 were engaged in sales, marketing, operations, and administrative positions.

The Compensation Committee of our Board of Directors assists in the oversight of policies and strategies relating to human capital management.

To execute our business strategy successfully, we must recruit, develop, and retain the very best talent globally, including exceptional executives, scientists, engineers, and technical and non-technical staff.

As the demand for global technical talent continues to be high, we have grown our technical workforce and have been successful in attracting top talent to NVIDIA.

We have attracted talent worldwide through our strong employer brand and differentiated hiring strategies for college, professional, and leadership talent.

Our workforce is 82% technical and 51%

hold advanced degrees.

Additionally, we have increased our focus on diversity recruiting and we welcome employees of all backgrounds.

Our own employees also help to surface top talent, with over 41% of our new hires in fiscal year 2025 coming from employee referrals.

Development and Retention

To support employee development, we provide opportunities to learn on-the-job through training courses, targeted development programs, mentoring and peer coaching and ongoing feedback.

We constantly upgrade our learning offerings to ensure that our employees are exposed to the most current content and technologies available.

We offer tuition reimbursement programs to subsidize educational programs and advanced certifications and encourage internal job mobility.

We have also implemented specifically designed mentoring and development programs for women and employees from traditionally underrepresented groups to ensure widespread readiness for future advancement.

To evaluate employee sentiment and engagement, we use pulse surveys, a suggestion box, and an anonymous third-party platform.

We want NVIDIA to be a place where people can grow their careers over their lifetime and our employees tend to come and stay.

In fiscal year 2025, our overall turnover rate was 2.5%.

Compensation, Benefits, and Well-Being

Our compensation program rewards performance and is structured to encourage employees to invest in the Company’s future.

Employees receive equity, except where unavailable due to local regulations, that is tied to the value of our stock price and vests over time to retain employees while simultaneously aligning their interests with those of our shareholders.

We offer comprehensive benefits to support our employees’ and their families’ physical health, well-being, and financial health.

Programs include 401(k) programs in the U.S., statutory and supplemental pension programs outside the U.S., our employee stock purchase program, flexible work hours, and time off policies.

We evaluate our benefit offerings globally and aim to provide comparable support across the regions where we operate.

We offer tailored benefits based on the needs of our employees including continuing support for parents;

both new birth parents and those who wish to become parents.

Our support is enhanced during times of crisis, such as war or economic volatility, to take care of our existing team of world-class talent and their families.

Diversity, Inclusion and Belonging at NVIDIA

We believe that diverse teams fuel innovation, and we are committed to creating an inclusive culture that supports all our employees.

When recruiting new talent or developing our current employees, we strive to build a diverse talent pipeline that includes those underrepresented in the technology field, including women, Black/African American, and Hispanic/Latino candidates.

To this end, we:

• Hire, promote and compensate our employees based on merit;

• Partner with institutions and professional organizations serving historically underrepresented communities;

• Enlist dedicated recruiting teams to shepherd underrepresented candidates through the interview process and identify internal opportunities;

• Support the development and growth of women employees through programs aimed at building a pipeline of future leaders;

• Provide peer support and executive sponsors for our internal community resource groups;

• Provide training and education to managers and peers on fostering supportive environments that allow all our employees to do their best work;

• Track equity and parity in retention, promotions, pay, and employee engagement scores;

• Measure year over year progress and provide leadership visibility on talent and diversity efforts.

As of the end of fiscal year 2025, our global workforce was 78% male, 21% female, and 1% not declared, with 6% of our workforce in the United States composed of Black or African American and Hispanic or Latino employees.

Flexible Working Environment

We support a flexible work environment allowing us to recruit the very best employees, regardless of where they live.

This flexibility supports diverse hiring, retention of talent, including working parents and other caregivers and employee engagement, which we believe makes NVIDIA a great place to work.

We also provide company-wide 2-days off per quarter for employees to rest and recharge.

To execute our business strategy successfully, we focus on recruiting, developing, and retaining top global talent.

Within our workforce, more than 80 percent have technical roles and more than half of the workforce hold an advanced degree.

Our employees also help to surface top talent, with over 40 percent of our new hires in fiscal year 2026 coming from employee referrals.

In fiscal year 2026, our turnover rate was 3.7 percent.

We invest in employee development through on-the-job trainings and tuition reimbursement programs.

Our compensation and benefits are designed to reward performance and align employee interests with those of our shareholders through equity participation and comprehensive health and financial wellness programs.

We also utilize employee listening systems to gather feedback and maintain an inclusive culture where hiring and promotions are based on merit.

Information About Our Executive Officers

10 unchanged sentences

From 1983 to 1985, Mr.

Huang was a microprocessor designer for AMD, a semiconductor company.

Huang was a microprocessor designer for AMD, a semiconductor

Huang holds a B.S.E.E.

41 unchanged sentences

Compared sentence by sentence after normalising whitespace, quotation marks, case and digits, so re-formatting and restated figures do not read as changed language. Wording changes appear as one removal and one addition. The current filing and the prior one are authoritative.

What changed 10-K

Item 1. Business