# Advanced Micro Devices

Advanced Micro Devices, Inc. (AMD) is an American multinational semiconductor company headquartered in Santa Clara, California, that develops CPUs, GPUs, FPGAs, and SoCs for PCs, gaming, data centers, and embedded systems. It is a major competitor to Intel and NVIDIA in the AI accelerator market.

Advanced Micro Devices, Inc. (AMD) is an American multinational semiconductor company headquartered in Santa Clara, California. It designs and develops central processing units (CPUs), graphics processing units (GPUs), field-programmable gate arrays (FPGAs), system-on-chips (SoCs), and high-performance computing components. AMD serves a wide range of business and consumer markets, including personal computers (PCs), gaming, data centers, and embedded systems. The company is a primary competitor to [Intel](https://www.wikiprompt.org/wiki/intel) in the CPU market and to [NVIDIA](https://www.wikiprompt.org/wiki/nvidia) in the GPU and data center AI accelerator market.

AMD's main products include microprocessors, embedded processors, and graphics processors for servers, workstations, PCs, and embedded system applications, as well as chipsets for motherboards. The company has expanded into data centers, gaming, and high-performance computing. Its processors are used in a wide range of computing devices, including PCs, servers, laptops, and gaming consoles. Initially manufacturing its own processors, AMD outsourced manufacturing after spinning off [GlobalFoundries](https://www.wikiprompt.org/wiki/tsmc) in 2009. Through its acquisition of Xilinx in 2022, AMD offers FPGA products.

## History

### Foundational years
AMD was formally incorporated on May 1, 1969, by Jerry Sanders and seven colleagues from Fairchild Semiconductor: Ed Turney, John Carey, Sven Simonsen, Jack Gifford, Frank Botte, Jim Giles, and Larry Stenger. Sanders, an electrical engineer and former director of marketing at Fairchild, left to start his own semiconductor company, following the path of Robert Noyce and Gordon Moore, who had founded Intel in July 1968.

In September 1969, AMD moved to a new headquarters at 901 Thompson Place in Sunnyvale, California, which housed a wafer production facility. To secure a customer base, AMD initially acted as a second-source supplier for microchips designed by Fairchild and National Semiconductor. The company guaranteed quality control to United States Military Standard, an advantage in the early computer industry where chip reliability was a concern.

In November 1969, AMD produced its first product, the Am9300, a 4-bit MSI shift register, which began selling in 1970. In 1970, AMD produced its first proprietary product, the Am2501 logic counter. Its bestselling product in 1971 was the Am2505, the fastest multiplier available. That year, AMD entered the RAM chip market with the Am3101, a 64-bit bipolar RAM, and annual sales reached US$4.6 million.

AMD went public in September 1972. By 1975, the company was producing 212 products, 49 of which were proprietary, including the Am9102 static RAM and low-power Schottky MSI circuits. In 1975, AMD entered the microprocessor market with the Am9080, a reverse-engineered clone of the Intel 8080, and the Am2900 bit-slice microprocessor family. When Intel began installing microcode in its microprocessors in 1976, it entered a cross-licensing agreement with AMD, effective October 1976.

In 1977, AMD entered a joint venture with Siemens, which purchased 20% of AMD's stock. The two companies established Advanced Micro Computers (AMC) to develop microcomputers based on AMD's second-source Zilog Z8000 microprocessors. AMD bought out Siemens' stake in the American division in 1979 and closed Advanced Micro Computers in late 1981 after shifting focus to second-source Intel x86 microprocessors.

Total sales in fiscal year 1978 topped $100 million, and in 1979, AMD debuted on the New York Stock Exchange. Production began at a new semiconductor fabrication plant in Austin, Texas, in 1979, with overseas assembly facilities in Penang and Manila. In 1980, AMD began supplying semiconductor products for telecommunications.

### Intel partnership and legal battles
Intel introduced the first x86 microprocessors in 1978. In 1981, IBM created its personal computer and wanted Intel's x86 processors, but only under the condition that Intel provide a second-source manufacturer. Intel and AMD entered a 10-year technology exchange agreement, allowing AMD to build legal clones of Intel processors, including the x86 family from 1981.

In the mid-1980s, Intel refused to share the blueprint of its i386 processor with AMD, leading to a seven-year legal battle. The dispute ended in AMD's favor in 1994, but required AMD to develop its own x86 microprocessors in-house. This legal history shaped AMD's independent x86 architecture development.

### Growth and challenges
In the early 2000s, AMD experienced significant growth, partly due to its strong position in the PC market and its Athlon and Opteron processors. The Opteron, introduced in 2003, was the first x86 processor with 64-bit extensions, giving AMD a competitive edge in servers.

AMD faced challenges in the late 2000s and early 2010s, struggling to compete with Intel. The company's acquisition of ATI Technologies in 2006 for $5.4 billion brought GPU capabilities but also financial strain. In 2009, AMD spun off its manufacturing operations into GlobalFoundries, transitioning to a fabless model.

### Ryzen and resurgence
In the late 2010s, AMD regained market share through a penetration pricing strategy for its Ryzen processors, offering competitive performance at lower costs than Intel microprocessors. The Ryzen line, based on the Zen microarchitecture, launched in 2017 and was well received in the consumer and data center markets. In 2022, AMD surpassed Intel by market capitalization for the first time.

## AI accelerators and ROCm

AMD has developed AI accelerators, notably the MI300X, to compete with NVIDIA in data center AI. The MI300X, announced in 2023, is a GPU designed for [large language models](https://www.wikiprompt.org/wiki/large-language-model) and other [AI](https://www.wikiprompt.org/wiki/artificial-intelligence) workloads. It features high-bandwidth memory (HBM3) and is optimized for [machine learning](https://www.wikiprompt.org/wiki/machine-learning) inference and training.

AMD's software stack for AI is called ROCm (Radeon Open Compute). ROCm provides a programming framework and libraries for GPU computing, supporting [deep learning](https://www.wikiprompt.org/wiki/deep-learning) frameworks such as PyTorch and TensorFlow. While ROCm has historically lagged NVIDIA's CUDA in ecosystem maturity, AMD has invested in improving compatibility and performance.

The MI300X competes directly with NVIDIA's A100 and H100 GPUs in data center deployments. AMD has positioned its accelerators as a cost-effective alternative, with a focus on open standards and multi-vendor support. As of 2025, AMD's AI accelerators are used in cloud services and enterprise data centers, though NVIDIA retains a dominant market share.

## Products and markets

AMD's product portfolio includes Ryzen processors for consumer PCs, EPYC processors for servers, Radeon GPUs for gaming and professional visualization, and Instinct accelerators for data center AI. The company also offers embedded processors and FPGAs through its Xilinx subsidiary.

In the gaming market, AMD supplies custom chips for gaming consoles, including the [Sony](https://www.wikiprompt.org/wiki/sony-ai) PlayStation and Microsoft Xbox. AMD's semi-custom business unit designs SoCs for these consoles, which have been a significant revenue source.

In data centers, AMD's EPYC processors have gained market share against Intel's Xeon line, particularly in cloud computing. The company's AI accelerators are used in [Amazon Web Services](https://www.wikiprompt.org/wiki/amazon-web-services), [Microsoft Azure](https://www.wikiprompt.org/wiki/azure), and [Google Cloud](https://www.wikiprompt.org/wiki/google-cloud) instances, among others.

## Manufacturing and technology

AMD outsources its manufacturing to foundries such as [TSMC](https://www.wikiprompt.org/wiki/tsmc) and GlobalFoundries. The company designs chips using advanced process nodes, including 5-nanometer and 3-nanometer technologies. AMD's chiplet design approach, introduced with the Zen 2 architecture, allows for modular integration of compute dies and I/O dies, improving yields and flexibility.

AMD's acquisition of Xilinx in 2022 expanded its capabilities in FPGAs and adaptive computing. Xilinx's Versal platform integrates FPGA fabric with AI engines, targeting applications in [machine learning](https://www.wikiprompt.org/wiki/machine-learning) inference and signal processing.

## Competitive landscape

AMD competes with Intel in CPUs, NVIDIA in GPUs and AI accelerators, and [Qualcomm](https://www.wikiprompt.org/wiki/qualcomm) in embedded and low-power processors. In the AI accelerator market, AMD faces competition from NVIDIA's CUDA ecosystem, as well as specialized startups like [Groq](https://www.wikiprompt.org/wiki/groq) and [SambaNova](https://www.wikiprompt.org/wiki/samba-nova). AMD's strategy emphasizes open standards, such as the ROCm software stack, and competitive pricing.

As of 2025, AMD's market position in AI is growing but still secondary to NVIDIA. The company's MI300X has been adopted by several cloud providers and enterprises, but NVIDIA's dominance in training and inference remains strong. AMD continues to invest in software and hardware to close the gap.

## See also

- [Intel](https://www.wikiprompt.org/wiki/intel)
- [NVIDIA](https://www.wikiprompt.org/wiki/nvidia)
- [TSMC](https://www.wikiprompt.org/wiki/tsmc)
- [Artificial intelligence](https://www.wikiprompt.org/wiki/artificial-intelligence)
- [Machine learning](https://www.wikiprompt.org/wiki/machine-learning)

---
Source: https://www.wikiprompt.org/wiki/advanced-micro-devices
License: CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/)
Last updated: 2026-09-13T03:56:17.414842+00:00