IoT, 5G mobile communication, autonomous driving vehicles and other emerging innovative industries highlight the growing importance of data storage and processing. At the core of such capabilities lie semiconductor memory devices. Korea has positioned itself as the unrivaled No. 1 leader in the semiconductor memory market over the years, supplying exceptional DRAM and other memory products. The development of semiconductor memory devices, however, has recently faced limitations in terms of integration density. Meanwhile, research on wire connections made among devices is gaining importance.

The Technology Enablement of Advanced MOS structure (TEAM) led by professor Rock-Hyun Baek at the Department of Electrical Engineering, POSTECH, is conducting research on the latest technologies that hold the greatest potential for commercialization among technologies currently sought after in the semiconductor industry. The lab consists of the ‘Logic Group’ which studies semiconductors used for mobile APs or personal computer CPUs, the ‘TSR Group’ which engineers and measures actual devices, the ‘Memory Group’ which probes into NAND flash memory used for Solid State Drives (SSDs), and the ‘ML Group’ which combines the latest AI technology of machine learning with semiconductor research.

The Logic Group, responsible for the study of semiconductor devices, is engaged in general research on transistors that form the basis for semiconductor devices. Over the past decades, transistors have evolved such in a way as to increase integration density while reducing their size. Samsung Electronics of Korea and TSMC of Taiwan, the undisputed leaders in the semiconductor memory segment, employ 5-7 nm processes (1 nanometer = one billionth of a meter).

Researchers at the TEAM leverage simulation technology to manufacture transistors that are even smaller than those made by these companies. They also evaluate their performance and present future guidelines for the semiconductor industry to follow. University laboratories usually face difficulties in building manufacturing equipment on the scale often achieved by private sector businesses, and instead perform research primarily through simulations. As devices continue to become more compact, the shape is also changing from somewhat flat designs to three-dimensional structures. TEAM researchers are interested in how these structural changes will affect performance.

In particular, work is underway to measure and analyze ‘Nanowire FETs’ or ‘NAND flash’ devices directly donated by Samsung Electronics or SK Hynix, who have yet to commercialize these devices. The TEAM plans to study specific performance indicators displayed during operation at low temperatures as well as the metal lines that connect the transistors. Another research topic concerns semiconductor technology enabled by machine learning. These endeavors will surely help nurture a professional semiconductor workforce armed with a significant global edge on the market.