How does Samsung plan to use AI to develop semiconductor materials faster?

Samsung aims to use AI to identify promising material candidates and optimal conditions while reducing the number of experiments required by more than 80 percent. Choi said the conventional approach relies on researchers’ experience and intuition, and does not fully use data from failed experiments.

Why is it difficult for Samsung researchers to find new semiconductor materials?

A seven-layer structure can involve more than 100 million possible material combinations, while researchers can run only about 50 experiments each month. Choi said identifying candidate materials through conventional methods can take months or even years.

Could Samsung’s AI approach to semiconductor materials be used in other industries?

Samsung expects the approach could lower development costs and expand AI use into other industries, including the development of battery materials.

Samsung aims to cut chip material development time using AI

Listen

Summary

Samsung Electronics said Monday it aims to use AI to reduce the time needed to develop new semiconductor materials from five years to two. At a Korea Industry Association Forum, Samsung Advanced Institute of Technology official Choi Young-sang said the company also seeks to identify more than five times as many promising material candidates. Samsung aims to cut the experiments needed to find optimal conditions by more than 80 percent and expects the approach could also be used for battery materials.

Key Facts

A seven-layer semiconductor structure can have more than 100 million possible material combinations, Choi said.
Researchers can conduct only about 50 experiments a month.
Choi said conventional research can take months or years to identify candidate materials and has not fully used data from failed experiments.
Samsung announced its goal at a forum hosted by the Korea Industry Association Forum on AI use in industrial research and development.

By Yonhap

Published Oct 26, 2026 2:40 pm KST

Samsung Electronics’ headquarters in Suwon, Yeongtong District, Gyeonggi Province, May 27 / Korea Times photo by Jeong Da-bin

Samsung Electronics Co. said Monday it aims to use artificial intelligence (AI) to cut the time needed to develop new semiconductor materials by three years.

The company announced the goal at a forum hosted by the Korea Industry Association Forum, where participants discussed the use of AI in industrial research and development (R&D).

"We aim to shorten the development period for new semiconductor materials from five years to two years using AI and increase the number of promising material candidates identified by more than fivefold," Choi Young-sang, an official at the Samsung Advanced Institute of Technology, the company's R&D center, said at the forum.

Choi said semiconductor materials consist of multiple thin-film layers, creating more than 100 million possible material combinations for a seven-layer structure. However, researchers can conduct only about 50 experiments a month.

"Under the conventional approach, which relies on researchers' experience and intuition, it can take months or even years to identify candidate materials. Data from failed experiments also has not been fully utilized," Choi pointed out.

Samsung aims to reduce the number of experiments needed to identify optimal conditions by more than 80 percent, Choi said.

Samsung also expects such an approach could lower development costs and expand the use of AI into other industries, including battery materials.

Explore More