Samsung Electronics Foundry to Quadruple Customers by 2029, Expand HPC Share to 66%

Roadmap Revealed at Korea Premium Week 2026

Accelerating Toward 1.4nm Mass Production and Taylor Fab Operation by 2029

Samsung Electronics has announced that by 2029, it will increase the number of its foundry (semiconductor contract manufacturing) customers to approximately four times that of when the standalone business division was first established. The company is also expanding its business domain into areas such as artificial intelligence (AI) and high-performance computing (HPC).

At the Korea Premium Week 2026, the largest capital market investor relations (IR) event in Korea, held at The Grand Lotte Seoul in Jung-gu, Seoul, on September 29, Noh Mi-jeong, executive director of the Foundry Business Division at Samsung Electronics, stated, "Currently, the number of foundry customers has increased about threefold compared to 2017, when the division became independent," adding, "By 2029, we expect customer numbers to grow by around four times compared to the time of establishment."

Reuters Yonhap News

Samsung Electronics plans to restructure its foundry business portfolio, which has been heavily concentrated on mobile smartphone semiconductors, to focus on AI and HPC. The share of HPC in total revenue, which stood at 5% in 2017, has now expanded dramatically to 28%, and it is projected to rise further to 66% by 2029.

Noh also stated, "By 2029, the majority of our customers will be in the HPC sector, and HPC will account for well over half of total sales." HPC refers to advanced computing semiconductor systems that process massive amounts of data at ultra-high speeds, such as for supercomputers and large-scale AI model training and inference.

Samsung Electronics is aiming to start mass production using the third-generation 2-nanometer (nm; 1 nm = one billionth of a meter) process in 2028, and further shift to mass production using the 1.4-nanometer process, known as the dream-class ultra-fine process, in 2029. Additionally, the first phase of its Taylor, Texas fab in the United States will begin initial operation at the end of this year, with the goal of full mass production by 2030. The company also plans to begin construction of a second fab within this year, ensuring a proactive response to next-generation AI semiconductor demand.

This content was produced with the assistance of AI translation services.