Samsung to Build Second Texas Semiconductor Plant for AMD and Nvidia Chip Production
Samsung Electronics is preparing to build a second semiconductor manufacturing facility in Taylor, Texas, aimed at producing advanced chip components for leading U.S. technology companies, including AMD and Nvidia. The proposed plant will manufacture base dies for high-bandwidth memory (HBM) chips, a critical element in artificial intelligence (AI) processors and high-performance computing systems.
The project represents a major expansion of Samsung's semiconductor manufacturing presence in the United States at a time when growing demand for AI infrastructure is driving significant investment in advanced chip production.
Construction of the new facility is expected to begin in November 2026. According to reports, the plant's production capacity has already been fully committed, with AMD accounting for the majority of the orders and Nvidia securing a portion of the remaining output.
Spanning approximately 2.7 million square feet, the facility will be around 50% larger than Samsung's existing Taylor manufacturing plant. The site is expected to utilize advanced 2-nanometer process technology, positioning it among the most sophisticated semiconductor manufacturing operations in the country.
Facility to Produce Critical HBM Components
The new Taylor facility will focus on producing base dies used in high-bandwidth memory chips. These components form the foundational logic layer within HBM stacks and are essential for managing communication between memory layers and processing systems.
HBM technology enables the rapid transfer of massive volumes of data between memory and processors, making it indispensable for AI accelerators, machine learning workloads, generative AI applications, and large-scale data centers. As demand for these technologies continues to grow, the need for advanced HBM components has become increasingly important across the semiconductor industry.
By dedicating the facility to HBM-related production, Samsung is positioning itself to capitalize on the expanding market for advanced memory technologies while strengthening relationships with major U.S. chip designers seeking reliable access to leading-edge manufacturing capacity.
Construction and Production Timeline
Equipment installation at the new plant is scheduled to begin in February 2029, with mass production expected to start in 2030.
The facility will expand Samsung's existing semiconductor operations in Williamson County, where the company already operates a manufacturing plant as part of its broader U.S. investment strategy. Samsung's first Taylor facility reportedly includes a manufacturing agreement with Tesla.
At approximately 2.7 million square feet, the second facility will significantly exceed the size of the company's original Taylor site, further reinforcing the region's importance within Samsung's global manufacturing network.
AI Boom Driving Semiconductor Investment
Samsung's planned expansion comes as semiconductor manufacturers worldwide increase investments in advanced fabrication technologies to meet surging demand from AI, cloud computing, and high-performance computing markets. Companies such as AMD and Nvidia continue to expand their AI hardware portfolios, increasing the need for secure and scalable access to advanced manufacturing and memory solutions.
For Samsung, the new Texas facility represents an opportunity to strengthen its foundry business and compete more aggressively in the advanced semiconductor manufacturing sector. Producing critical components for major chip designers could further enhance the company's role in the supply chains powering next-generation AI processors.
The investment also highlights the growing importance of Central Texas as a hub for semiconductor manufacturing. Supported by infrastructure development, a skilled workforce, and proximity to major technology companies, the Taylor region continues to attract significant investment and contribute to the expansion of the U.S. technology manufacturing ecosystem.
Project News
