Semiconductors are the foundation of artificial intelligence (AI), a technology that is transforming our economy and society, making entire industries more productive and innovative, and driving major scientific breakthroughs.
Today鈥檚 AI systems are built on decades of innovation across the entire semiconductor ecosystem. As chip technologies continue to advance, AI will become more capable, energy-efficient, and cost effective. Stronger AI, in turn, will help improve chip design, optimize semiconductor manufacturing, and drive more demand for the broad range of chips that enable AI.
Key Takeaways:
Chips provide the base hardware layer underpinning modern AI systems and comprise a significant portion of the overall value in a modern AI server:
鈥� A single AI server rack contains over 4,500 packaged chips, comprised of approximately 20,000 individual dies 鈥� i.e., unique integrated circuits.
鈥� Semiconductors account for more than 95% of a leading AI server rack鈥檚 content value and more than 50% of the total capital expenditures required for building and operating an AI data center.
鈥� The stability of the entire AI data server rests on a long tail of lower-cost foundational chips 鈥� including compound semiconductors made from more than one material 鈥� that account roughly 10% of the chip content by value, but the majority of chips in a server by volume
To run complex AI training and inference workloads, today鈥檚 AI data centers need huge amounts of compute, storage and memory bandwidth, power distribution, and networking capabilities 鈥� all provided by the full stack of chip technologies. Each of these chip technologies is essential to driving America鈥檚 AI buildout, while supply disruptions in any of these areas could risk hampering this buildout. Chips in AI data centers include:
鈥� Advanced logic chips, such as AI accelerators, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), central processing units (CPUs), data processing units (DPUs), and networking chips.
鈥� Memory, such as high-bandwidth memory (HBM), both dynamic and static random-access memory (DRAM and SRAM), and non-volatile flash memory (NAND).
鈥� Analog and foundational chips, such as power chips, transceivers, controllers, and sensors.
In a positive feedback loop, advances in AI drive demand for improved semiconductor performance and efficiency, while improvements in semiconductor technology enable more powerful and advanced AI systems.
鈥� To meet global demand for new AI applications, government and industry will invest over $4 trillion in new data center infrastructure through 2028, of which up to $2.8 trillion will be spent on semiconductors.
鈥� Annual revenue for semiconductors deployed in AI data centers could reach over $1.2 trillion by 2028, a nearly tenfold increase over four years.
鈥� The AI data center market is experiencing unprecedented growth, projected at an 88.8% Compound Annual Growth Rate (CAGR) from 2022 to 2028. While initial momentum was driven by the rapid adoption of Generative AI, sustained demand remains robust, with a projected 56.3% CAGR from 2025 through 2028.
The entire semiconductor supply chain enables the buildout of AI infrastructure. Without semiconductors, there is no AI. To lead in this transformational technology, government and industry must work together to advance policies to accelerate growth and innovation across the full spectrum of chip technologies and work closely with global partners to build strong and resilient supply chains.
Policy Recommendations:
To promote continued American leadership in semiconductor and AI technologies, policymakers should adopt pro-investment, pro-innovation, and pro-growth policies to strengthen semiconductor research, design, and manufacturing:
