High Bandwidth Memory Challenges in the AI Era

10:45 am - 11:10 am

Since its inception, the memory semiconductor industry has continuously evolved toward achieving higher density and higher speeds. Driven by the rapid growth of the AI market, these requirements have escalated further, making High Bandwidth Memory (HBM) one of the most important components of AI infrastructure.

While increasing capacity and operating speed remain primary objectives for HBM, achieving them presents significant technical hurdles. These include the implementation of high-stack packaging technologies, as well as managing increased power consumption and high-thermal dissipation resulting from high-speed operations.

Furthermore, with the generation transition cycle of HBM shortening significantly compared to the past, securing proactive core unit technologies has become paramount. This presentation will examine the key challenges facing HBM development and discuss technical solutions as well as future strategic directions to overcome these challenges.

Key Technologies Covered

High Bandwidth Memory
Memory for AI Infrastructure

Featured Speakers

Mr. SEUNGTAEK YANG

Director, SK Hynix

Seungtaek Yang is a highly experienced semiconductor packaging expert with over two decades of industry experience. As of 2026, he is based at the company's Taiwan subsidiary, where he leads strategic collaborations and HBM-focused partnerships with major Taiwanese tech companies.

He joined Hynix in 2003, beginning his career as a Wafer-Level Packaging (WLP) Process Development Engineer. After building a strong foundation in advanced packaging processes, he transitioned to WLP Product Development in 2006. Since 2010, he has played a pivotal role in the development of cutting-edge HBM (High Bandwidth Memory) and 3DS products by pioneering work on Through-Silicon Via (TSV) technology. He earned a B.S. in Metallurgical Engineering from Ulsan University in 2001 and an M.S. in Materials Science and Engineering from Hanyang University in 2003.