AMD

AMD, short for Advanced Micro Devices, is a global semiconductor and computing company that designs CPUs, GPUs, AI accelerators, adaptive computing products, networking technology, embedded processors, and supporting software. Best known to consumers for Ryzen processors and Radeon graphics, AMD now operates across PCs, gaming, cloud computing, enterprise servers, artificial intelligence, high-performance computing, networking, and embedded systems.

Modern AMD is broader than a conventional PC-chip company. Its portfolio combines processor design, graphics, AI acceleration, adaptive computing, data-center networking, software, and increasingly rack-scale infrastructure.

Last updated: August 2026

About AMD

Advanced Micro Devices, Inc. is an American semiconductor company headquartered in Santa Clara, California. Founded in 1969, AMD develops computing technologies for consumer, enterprise, cloud, embedded, and data-center markets.

The company designs its own processor architectures and semiconductor products but relies on external manufacturing partners to fabricate its chips. This fabless operating model allows AMD to focus heavily on architecture, product design, intellectual property, and software rather than owning leading-edge semiconductor fabrication plants.

AMD operates through three reportable business segments: Data Center, Client and Gaming, and Embedded.

Its best-known brands include Ryzen processors, Radeon graphics, EPYC server CPUs, Instinct AI accelerators, Pensando networking products, and adaptive computing technologies inherited and expanded through its acquisition of Xilinx.

What Does AMD Do?

AMD designs computing technology for several major markets.

Its Client and Gaming business supplies processors and graphics technologies used in desktop computers, notebooks, workstations, gaming systems, and semi-custom products.

Its Data Center business develops server CPUs, AI accelerators, GPUs, networking products, DPUs, AI networking technology, FPGAs, and adaptive computing products for cloud providers, enterprises, research organizations, and large-scale AI infrastructure.

Its Embedded business supplies processors, FPGAs, adaptive SoCs, and related technologies for specialized systems such as industrial equipment, communications infrastructure, embedded computing, automotive applications, healthcare systems, aerospace, and edge devices.

AMD also develops software platforms and developer tools. ROCm, for example, supports GPU computing and AI workloads on compatible AMD hardware.

The result is a business that now spans much more of the computing stack than it did during the period when AMD was mainly known as an x86 processor competitor.

AMD Business Model

AMD primarily operates through a fabless semiconductor business model.

The company develops processor architectures, chip designs, intellectual property, software, and computing platforms internally, while external foundries manufacture the silicon wafers.

For leading-edge microprocessors and GPUs, AMD relies heavily on TSMC. It also works with GlobalFoundries and other manufacturing partners for selected products and process technologies.

A simplified version of the business model looks like this:

AMD architecture and product design → external semiconductor fabrication → assembly, testing and packaging → sale through OEMs, cloud providers, enterprises, distributors, retailers, and other customers

This structure differs from semiconductor companies that both design and manufacture substantial portions of their own chips.

AMD's model reduces the need to fund and operate cutting-edge fabrication plants, but it also creates dependencies on manufacturing partners, available production capacity, process technology, yields, packaging, and supply-chain execution.

The business model has also become more integrated at the product level.

AMD increasingly combines:

• CPUs

• GPUs and AI accelerators

• adaptive computing

• networking

• software

• system-level design

This is particularly important in data centers and artificial intelligence, where customers increasingly evaluate complete computing platforms rather than individual chips in isolation.

How Does AMD Make Money?

AMD generates revenue through three reportable segments.

AMD reported $34.6 billion in net revenue for fiscal 2025, representing a 34% increase from 2024.

Data Center

Data Center has become one of AMD's most strategically important businesses.

The segment includes technologies such as:

• EPYC server processors

• Instinct AI accelerators

• Pensando networking products

• data-center GPUs

• DPUs

• AI network interface technologies

• FPGAs

• adaptive SoCs

The growth of cloud computing and AI infrastructure has significantly increased the importance of this segment.

AMD reported $6.7 billion in Data Center revenue in Q2 2026, representing 58% of the company's total quarterly revenue.

Client and Gaming

Client and Gaming includes AMD's PC and gaming technologies.

Important products include:

• Ryzen desktop and notebook processors

• Ryzen AI processors

• Radeon graphics

• chipsets

• semi-custom gaming technologies

The segment generated approximately $14.6 billion in fiscal 2025 revenue.

Embedded

AMD's Embedded segment serves markets where computing products are integrated into specialized systems rather than conventional PCs.

Applications can include industrial systems, communications infrastructure, embedded devices, healthcare equipment, aerospace systems, and edge computing.

Embedded generated approximately $3.5 billion in fiscal 2025 revenue.

Products and Services

AMD's product portfolio covers consumer computing, enterprise systems, data centers, AI, graphics, embedded systems, networking, and developer software.

AMD Ryzen Processors

Ryzen is AMD's main processor family for desktop and notebook computers.

The wider Ryzen portfolio includes products for mainstream PCs, business systems, high-performance desktops, and AI-capable computers.

Ryzen PRO products target commercial systems, while Ryzen Threadripper serves demanding workstation workloads.

AMD EPYC Server Processors

EPYC is AMD's server CPU family.

EPYC processors are designed for cloud computing, enterprise servers, high-performance computing, supercomputing, and AI infrastructure.

The expansion of EPYC has played an important role in AMD's growth within the data-center market.

AMD Radeon Graphics

Radeon is AMD's main consumer graphics brand.

Radeon GPUs are used primarily for gaming and other graphics-intensive workloads, while Radeon PRO products target professional visualization and workstation applications.

AMD entered the graphics-processor business through its acquisition of ATI Technologies in 2006.

AMD Instinct AI Accelerators

AMD Instinct accelerators target artificial intelligence and high-performance computing.

They are designed for workloads such as AI training, inference, scientific computing, and large-scale accelerated computing.

Instinct forms a central part of AMD's effort to compete in AI infrastructure.

Adaptive Computing and Xilinx Products

AMD's acquisition of Xilinx significantly expanded its adaptive and embedded computing portfolio.

The company now offers FPGA and adaptive computing families including:

• Versal

• Virtex

• Kintex

• Artix

• Spartan

• Zynq

These technologies serve specialized workloads in markets such as communications, industrial computing, embedded systems, data centers, and edge infrastructure.

AMD Pensando Networking

Pensando products extend AMD into data-center networking and infrastructure processing.

Pensando DPUs can offload infrastructure workloads from host CPUs, while other networking technologies support data movement across increasingly large computing clusters.

Networking has become particularly important as AI infrastructure scales to systems containing large numbers of accelerators.

ROCm Software

ROCm is AMD's software platform for GPU computing, artificial intelligence, and high-performance computing.

It includes tools, libraries, runtimes, compilers, and framework integrations for developing and deploying workloads on supported AMD hardware.

ROCm is strategically important because AI competition depends not only on accelerator hardware but also on the software environment available to developers and customers.

Customers and Markets

AMD serves both consumer and business markets.

Consumer and PC Customers

Consumers encounter AMD primarily through:

• Ryzen desktop processors

• Ryzen laptop processors

• Ryzen AI systems

• Radeon graphics

• gaming products

AMD products are sold through PC manufacturers, system builders, distributors, retailers, and other channel partners.

Cloud and Data Center Customers

AMD supplies technology for:

• cloud computing

• hyperscale infrastructure

• enterprise servers

• AI infrastructure

• high-performance computing

• supercomputing

Products in these environments can include EPYC CPUs, Instinct accelerators, Pensando networking, adaptive computing products, and AMD software.

Embedded and Industrial Markets

AMD also serves industries where processors and adaptive computing technologies are incorporated into specialized equipment.

Relevant markets include:

• industrial systems

• telecommunications

• automotive

• healthcare technology

• aerospace and defense

• test and measurement

• broadcasting

• edge computing

This market diversity reduces AMD's dependence on a single type of computing device, although individual segments still move through different demand and product cycles.

Competitors

AMD competes with different companies in different parts of its business.

CPU Competition

Intel is AMD's primary competitor in CPUs and APUs.

The two companies overlap significantly in desktop, notebook, commercial PC, and server processors.

Graphics and AI Competition

Nvidia is AMD's principal competitor in discrete graphics.

The companies also compete heavily in accelerated computing and artificial intelligence through products such as AMD Instinct and Nvidia's data-center accelerator platforms.

Data Center Competition

Competition becomes more complex in data centers because customers increasingly evaluate complete computing environments.

AMD now competes through combinations of:

• EPYC CPUs

• Instinct accelerators

• Pensando networking

• ROCm software

• adaptive computing

• system-level infrastructure

Detailed AMD-vs-Intel or AMD-vs-Nvidia product comparisons are better treated separately because product generations, benchmarks, pricing, and software compatibility change frequently.

Company History

AMD was founded in 1969, but several strategic turning points explain how it became the company it is today.

The ATI acquisition gave AMD the graphics capabilities that became central to Radeon and helped establish a long-term combination of CPU and GPU technologies.

The 2009 manufacturing restructuring changed AMD's operating model by moving major fabrication assets into the business that became GlobalFoundries.

The introduction of Ryzen and EPYC in 2017 marked another major transition. AMD's Zen architecture gave the company a platform that could compete across both consumer and server computing.

The acquisitions of Xilinx and Pensando in 2022 then extended AMD beyond CPUs and GPUs into adaptive computing and networking.

AMD completed its acquisition of ZT Systems in March 2025 to strengthen its AI system-design capabilities. It subsequently sold ZT Systems' data-center infrastructure manufacturing operation to Sanmina while retaining selected design and customer-enablement capabilities.

Leadership

Dr. Lisa Su — Chair and CEO

Dr. Lisa Su serves as Chair and Chief Executive Officer of AMD.

She joined AMD in 2012 and became President and CEO in October 2014 before later becoming Chair of the Board.

Her leadership period coincided with several of AMD's most important strategic developments, including the expansion of the Zen architecture, Ryzen processors, EPYC server CPUs, data-center computing, adaptive computing, and AI infrastructure.

AMD's current direction reflects a broader focus on high-performance and adaptive computing rather than competing mainly through PC processors.

Leadership responsibilities across AMD extend across product engineering, data centers, client computing, gaming, embedded systems, software, operations, finance, and corporate functions.

Ownership, Subsidiaries and Corporate Structure

Advanced Micro Devices, Inc. is a publicly traded company.

Its common stock trades on the Nasdaq Global Select Market under the symbol AMD.

AMD does not operate under a corporate parent company.

Its main reportable business structure currently consists of:

• Data Center

• Client and Gaming

• Embedded

Major acquisitions have been integrated into this broader corporate structure.

Xilinx significantly expanded AMD's adaptive computing and embedded portfolio, while Pensando strengthened its networking capabilities.

ZT Systems added AI infrastructure design capabilities. AMD subsequently divested ZT Systems' manufacturing operation to Sanmina while retaining selected design expertise and intellectual property.

The company's corporate structure therefore increasingly reflects an integrated portfolio rather than a collection of completely separate acquired businesses.

AMD Strategy

AMD's current strategy centers on expanding its role across high-performance computing, artificial intelligence, data centers, client systems, and adaptive computing.

[H3] Expanding the AI Computing Stack

One of AMD's most important strategic priorities is AI infrastructure.

Instead of competing only through an individual accelerator, AMD increasingly combines:

EPYC CPUs + Instinct accelerators + Pensando networking + ROCm software + system-level design

This allows AMD to address a larger portion of the infrastructure required for AI training, inference, and large-scale accelerated computing.

Growing Data Center Revenue

The rising contribution of Data Center revenue shows how important cloud computing and AI have become to AMD.

Data Center represented 58% of company revenue in Q2 2026.

That changes AMD's strategic profile. Consumer computing remains important, but future growth increasingly depends on enterprise, cloud, AI, and hyperscale infrastructure.

Strengthening Software

ROCm is another strategic priority.

Competitive AI hardware needs a usable developer ecosystem. AMD therefore continues investing in software tools, framework support, libraries, and developer compatibility alongside its accelerator roadmap.

Expanding Rack-Scale Infrastructure

AMD is moving beyond individual semiconductor components toward more integrated AI systems.

Its work around rack-scale infrastructure brings together CPUs, accelerators, networking, and software into larger platforms designed for data-center deployment.

Maintaining a Broad Computing Portfolio

AMD is not abandoning PCs, gaming, or embedded systems as it expands into AI.

Ryzen, Radeon, embedded products, adaptive computing, and semi-custom technologies provide AMD with revenue exposure across several different computing markets.

Key Challenges and Risks

AMD's broader portfolio creates major opportunities, but it also exposes the company to several material risks.

Dependence on External Manufacturing

AMD relies on third-party foundries for semiconductor fabrication.

TSMC is particularly important for AMD's leading-edge microprocessors and GPUs.

Limited manufacturing capacity, poor yields, production delays, packaging constraints, or process-technology problems could therefore affect AMD's ability to supply products.

Intense Competition

AMD competes against large and technically sophisticated companies.

Intel remains a major processor competitor, while Nvidia is particularly strong in graphics and AI acceleration.

Competition also extends into networking, software, adaptive computing, and complete data-center platforms.

AI Software and Ecosystem Execution

AI infrastructure competition is not determined by accelerator hardware alone.

AMD must continue improving ROCm, framework compatibility, developer adoption, networking, and system-level integration if it wants customers to deploy its AI hardware at scale.

Product Execution

Semiconductor markets move through rapid product cycles.

Delays in architecture development, manufacturing readiness, software support, or platform deployment can weaken a product generation even when the underlying technology is competitive.

Supply-Chain Concentration

AMD's fabless model provides flexibility and reduces the need to own fabrication plants, but it also creates dependence on external suppliers.

Leading-edge manufacturing capacity is particularly difficult and expensive to replace quickly.

Key Facts About AMD

• AMD primarily operates as a fabless semiconductor designer, relying on external foundries rather than owning leading-edge fabrication plants.

• The company operates across processors, GPUs, AI accelerators, adaptive computing, networking, embedded systems, software, and AI infrastructure.

• AMD's acquisition of ATI Technologies established the foundation of its modern graphics business.

• The acquisition of Xilinx added a major adaptive computing and FPGA portfolio.

• Pensando expanded AMD into data-center networking and infrastructure processing.

• AMD's Data Center segment represented 58% of total company revenue in Q2 2026.

• AMD's AI strategy combines CPUs, accelerators, networking, software, and system design rather than focusing on a single AI chip.

• AMD relies heavily on TSMC for leading-edge microprocessor and GPU manufacturing.

• Dr. Lisa Su has led AMD as CEO since 2014.

AMD Case Studies

Detailed case studies can examine individual AMD strategic decisions separately from this company profile.

Potential areas for deeper analysis include:

• AMD's turnaround under the Zen architecture

• The strategic impact of the ATI acquisition

• AMD's acquisition of Xilinx

• AMD's expansion into AI accelerators

• AMD's data-center growth strategy

• AMD's fabless manufacturing model

• AMD vs Nvidia in AI infrastructure

No separate WeCaseStudy AMD case study is listed as published in the supplied material.

Frequently Asked Questions

What does AMD stand for?

AMD stands for Advanced Micro Devices. Its legal corporate name is Advanced Micro Devices, Inc.

What does AMD do?

AMD designs processors, graphics products, AI accelerators, adaptive computing technologies, embedded processors, networking products, and supporting software for PCs, gaming, cloud computing, data centers, AI, and specialized systems.

Does AMD manufacture its own chips?

AMD designs its semiconductor products but relies on third-party foundries for wafer fabrication. TSMC manufactures many of AMD's leading-edge microprocessor and GPU wafers.

Who is the CEO of AMD?

Dr. Lisa Su is AMD's Chair and Chief Executive Officer.

Is AMD a competitor to Intel and Nvidia?

Yes, but the overlap differs by market. Intel is a major competitor in CPUs and APUs, while Nvidia is AMD's principal competitor in discrete graphics and a major competitor in AI acceleration. AMD competes with both companies across parts of the data-center market.

What are AMD's main products?

AMD's major product families include Ryzen processors, Radeon graphics, EPYC server CPUs, Instinct AI accelerators, adaptive computing products such as Versal and FPGA families, Pensando networking products, and ROCm software.

How does AMD make money?

AMD generates revenue through three reportable segments: Data Center, Client and Gaming, and Embedded. Data Center has become particularly important as demand for cloud and AI infrastructure has expanded.

Who manufactures AMD processors?

AMD relies on outside semiconductor manufacturers. TSMC produces many of AMD's leading-edge microprocessor and GPU wafers, while GlobalFoundries and other suppliers manufacture selected products using other process technologies.

Resources:

• AMD 2025 Annual Report — Form 10-K

• AMD Q2 2026 Financial Results

• AMD Products and Technologies

• AMD Corporate Information

• Dr. Lisa Su — AMD Chair and CEO