JEJU, South Korea — As the semiconductor industry faces unprecedented technical hurdles, a clear consensus has emerged: the path forward lies not in chasing short-term product specifications, but in returning to the fundamental laws of physics.
At the 15th Semiconductor & Display Jeju Forum held on April 10, Yang Won-seok, an advisor to Wonik IPS, delivered a compelling keynote. He argued that to accurately predict the future roadmap of semiconductors, one must track the flow of physical constants—such as photons, atoms, and electric charges—rather than being obsessed with labels like "2-nanometer."
"Insight into the roadmap is not about guessing the next product; it is about understanding scaling rules and constants that never change," Yang stated. "In a rapidly shifting landscape, grasping 'common sense'—the immutable physical truths—is the key to forecasting the industry's trajectory."
The Shift from 'Area' to 'Performance'
Yang traced the evolution of semiconductors through the lens of PPAA (Power, Performance, Area, Availability). While the 1990s focused on cost reduction through Area scaling and the 2000s prioritized Power efficiency for the mobile era, we have now entered the age of Performance.
According to Yang, high-bandwidth memory (HBM) and advanced logic chips have pushed performance to the forefront of market value. "We have moved from an era of meeting 'specifications' to an era of delivering 'value,'" he noted. "Only by achieving overwhelming performance can a product survive in today’s competitive market."
Confronting Atomic Probability
The advisor also highlighted a critical shift in lithography. As processes shrink to the atomic level, the industry is transitioning from "deterministic engineering"—where results mirror the designer's intent—to a realm governed by statistical probability.
With photoresist (PR) layers becoming extremely thin, engineers must now manage errors as small as one or two silicon atoms. Yang warned that this challenge cannot be solved by mechanical advancement alone; it requires a deep integration of physical science to manage the inherent randomness of sub-nanoscale manufacturing.
The Future: Physical AI and Integrated Logistics
The forum also addressed the rise of "Physical AI" and the modernization of manufacturing facilities. Professor Jang Young-tae of KAIST pointed out a critical imbalance in the industry: while investment in production equipment is heavy, investment in logistics and data integration often lags.
"Just as the brain, nerves, and muscles in a human body move in unison, a factory's data, manufacturing IT, and logistics must operate as a single organic system," Jang emphasized. Innovative ideas, such as using autonomous drones for material transport within cleanrooms, were discussed as essential steps toward the next generation of semiconductor competitiveness.
[Copyright (c) Global Economic Times. All Rights Reserved.]
