====== AMD ======
**Advanced Micro Devices (AMD)** is a major semiconductor manufacturer that designs and produces processors, graphics processing units (GPUs), and system-on-chip (SoC) solutions for computing, gaming, data center, and embedded systems markets. Founded in 1969, AMD has emerged as a significant competitor in high-performance computing and GPU acceleration, though it operates in markets heavily dominated by established players like Intel and NVIDIA.

===== Company Overview =====
AMD designs and manufactures a broad range of semiconductor products across multiple market segments. The company operates a fabless business model, designing chips while outsourcing manufacturing to partners like Taiwan Semiconductor Manufacturing Company (TSMC). AMD's product portfolio includes **Ryzen processors** for consumer computing, **EPYC processors** for enterprise data centers, **Radeon GPUs** for graphics and compute acceleration, and **XILINX** programmable logic devices following its 2022 acquisition of Xilinx Inc. for approximately $49 billion (([[https://www.amd.com/en/press-releases/2022-02-14-amd-to-acquire-xilinx-expanding-high-performance-computing-and-adaptive-computing|AMD - AMD to Acquire Xilinx (2022]])).

===== GPU and Compute Acceleration =====
AMD's GPU business operates through its Radeon division, which produces discrete graphics cards and compute accelerators targeting gaming, professional visualization, and data center applications. The company's RDNA architecture family powers consumer Radeon RX series graphics cards, while the CDNA architecture targets compute-intensive workloads in AI and scientific computing. AMD's EPYC processors with integrated graphics and standalone accelerators compete in markets where GPU acceleration drives performance gains for machine learning inference and training workloads (([[https://www.amd.com/en/products/specifications/processors/server|AMD - EPYC Processor Specifications (2026]])).

In the competitive landscape of GPU manufacturing, AMD maintains a smaller market share compared to NVIDIA, which dominates the discrete GPU market particularly for AI and machine learning applications. The GPU accelerator market remains concentrated, with NVIDIA holding approximately 80-90% market share in data center GPU deployments as of 2025-2026 (([[https://www.semiengineering.com/ai-chips-reshape-computing/|Semiconductor Engineering - AI Chips Reshape Computing (2025]])).

===== Data Center and Enterprise Computing =====
AMD's EPYC server processors represent the company's primary challenge to Intel's historical dominance in enterprise computing. EPYC processors utilize zen-based CPU architectures with high core counts, advanced memory hierarchies, and support for modern instruction sets. Since EPYC's introduction in 2017, the processor line has captured significant market share in cloud computing infrastructure, with major cloud providers including Amazon Web Services, Google Cloud, and Microsoft Azure incorporating EPYC-based instances in their offerings (([[https://www.amd.com/en/press-releases/2023-01-10-amd-epyc-genoa|AMD - EPYC Genoa Launch Announcement (2023]])).

===== Market Position and Competitive Dynamics =====
AMD's competitive positioning reflects a multi-front strategy addressing different market segments. In consumer processors, AMD's Ryzen line competes directly with Intel's Core architecture. In data centers, EPYC competes with Intel's Xeon processors and increasingly with custom silicon developed by hyperscalers. In GPU acceleration, AMD's offerings address cost-sensitive segments and specific technical workloads, though NVIDIA's software ecosystem and market momentum present significant barriers to broader adoption (([[https://www.exponentialview.co/p/ev-570|Exponential View - Newsletter Issue 570 (2026]])).

The company's overall market capitalization and revenue scale substantially below NVIDIA's, reflecting differences in market penetration and strategic positioning. As of 2025-2026, AMD generates annual revenue exceeding $20 billion, with significant portions derived from processing unit sales rather than GPU accelerators.

===== See Also =====

  * [[amd_ryzen_ai_pro|AMD Ryzen AI Pro 400 Series]]
  * [[micron|Micron]]
  * [[wafer_scale_processors|Wafer-Scale Processors]]

===== References =====
  * https://www.amd.com/en/press-releases/2022-02-14-amd-to-acquire-xilinx-expanding-high-performance-computing-and-adaptive-computing
  * https://www.amd.com/en/products/specifications/processors/server
  * https://www.semiengineering.com/ai-chips-reshape-computing/
  * https://www.amd.com/en/press-releases/2023-01-10-amd-epyc-genoa
  * https://www.exponentialview.co/p/ev-570