www.koreatimes.co.kr, May. 29, 2025 –
Professor Kim Joung-ho of the Korea Advanced Institute of Science & Technology (KAIST), widely recognized as a leading expert in high-bandwidth memory (HBM) design, will present a 15-year roadmap for the next generation of artificial intelligence (AI) chips.
According to the university, KAIST Terabyte Interconnection and Package Laboratory (TERALAB) will host the Next-Generation HBM Roadmap (2025–2040) Technology Presentation online on June 11, offering a comprehensive look at the architecture, structure, performance and characteristics of next-generation HBM technologies from HBM4 to HBM8.
“The lab will host the event to present a strategic direction for Korea’s semiconductor industry amid intensifying global tech competition and to share a concrete roadmap for the development of next-generation HBM technology, a key component in AI semiconductors,” an official at the lab said.
Led by Kim, KAIST TERALAB has been at the forefront of HBM design technology for over two decades and has been directly involved in commercial HBM design since 2010.
The lab is widely recognized for its originality in research, in areas of HBM through-silicon via (TSV), interposers, signal integrity and power integrity design.
As of this month, 29 student researchers are conducting studies on the architecture, structure and performance of HBM4 to HBM8 that will span the next 15 years.
Currently, HBM3e is the most advanced commercially manufactured HBM chip. Memory chip makers such as Samsung Electronics and SK hynix are now preparing for the mass production of HBM4, which will offer higher bandwidth and improved energy efficiency as well as serve clients’ tailored needs.
KAIST TERALab is also working on the automation of HBM design using AI, by combining reinforcement learning and generative AI, to push the boundaries of high-bandwidth memory’s electrical and thermal optimization.
The upcoming presentation will not only cover the structure, performance and characteristics of each HBM generation but also introduce key enabling technologies aimed at dramatically boosting AI semiconductor computing speeds, KAIST said.
“The presentation is purposed at enhancing the industry’s understanding of the future of HBM technology, and to contribute to the sustainable growth of Korea’s semiconductor industry,” Kim said.
“The lab plans to share the presentation with major domestic chipmakers such as Samsung Electronics and SK hynix, and aims to hold similar presentations overseas, particularly in Silicon Valley.” Click here to read more