Samsung to Make AMD and Nvidia Memory Chip Parts at Second Texas Plant, Report Says

The newspaper, citing industry sources, said the plant has already secured orders for its entire planned output and that most of the volume is for AMD. Samsung declined to comment on the report, Reuters said.

The base die is the logic chip at the bottom of an HBM stack. It connects the stacked memory to the processor, and it matters more as AI accelerators from Nvidia and AMD need faster, denser memory.

What the Report Says

According to the report, the second Taylor facility will use 2-nanometre process technology and cover about 2.7 million square feet, about 1.5 times the size of the first plant on the site. The first plant has a contract with Tesla.

Construction is expected to start formally this month. Equipment installation would begin from February 2029, with mass production planned for 2030, the report said.

Why It Matters

If confirmed, the orders would give Samsung's US foundry business two large AI chip customers alongside Tesla. Samsung has spent several years trying to win leading-edge contract manufacturing business from rival TSMC. It would also put part of the supply chain for AI accelerator memory in the United States, where chipmakers face pressure to expand domestic production.

The report arrived during a broad sell-off in AI-related stocks. AMD shares were down 7.52% and Nvidia shares down 1.26% when Investing.com's version of the Reuters story was captured, though the figures did not show how much of the move was linked to the report.

Neither AMD nor Nvidia commented in the reports. Until Samsung or its customers confirm the orders, the details rest on the Korean newspaper's unnamed industry sources.