In a timed move, Samsung announced last Friday that it has begun shipping the industry's first 12-layer HBM4E samples to global customers. The rollout comes just days before Huang's upcoming visit to Seoul, following his appearance at the weeklong Computex trade show in Taipei, which wraps up on Friday.
The Nvidia CEO is expected to meet South Korean technology companies including Naver and LG, during his stay.
"I think there is something to the idea that this is a signal timed to Huang's scheduled visit to Seoul," Kim Duk-ki, a professor at Sejong University's Department of Semiconductor Systems Engineering, told AJP on Monday. "It is not bad from Nvidia's perspective at all, as it would encourage competition."
Kim said Nvidia stands to benefit from Samsung competing more aggressively with SK hynix, which has dominated HBM supply to the chip designer. As demand for AI infrastructure grows, global tech firms have been scrambling to secure sufficient memory supplies.
"Nvidia has every reason to want a competitive dynamic between suppliers," Kim said. "Stable volume is the priority, and competition helps ensure that."
The new 12-layer HBM4E component uses Samsung's sixth-generation 10-nanometer-class (1c) DRAM process paired with a 4-nanometer logic base die, boosting performance by more than 20 percent compared with previous-generation HBM4 chips.
"The growth in memory demand has been exponential and continues to accelerate," Kim said. "Samsung is utilizing this window to aggressively position itself for the next-generation supply chain."
Huang's Seoul itinerary, later this week or early next week, is also expected to include meetings with South Korean tech executives to discuss hardware supply chains and broader AI partnerships.
Copyright ⓒ Aju Press All rights reserved.
