Samsung Electronics is set to begin mass production of its next-generation high-bandwidth memory (HBM4) chips as early as next month, with plans to supply them to AI chip leader Nvidia, according to a source familiar with the matter.
The move signals a significant step in Samsung’s efforts to catch up with its domestic rival, SK Hynix, which has established itself as a primary supplier of advanced HBM chips for Nvidia’s AI accelerators. Samsung previously faced supply delays that impacted its earnings and share price in early 2023.
Market reaction to the news was immediate on Monday, with Samsung shares rising 2.2% while SK Hynix shares fell 2.9%.
The source declined to specify shipment volumes. Both Samsung and Nvidia declined to comment on the report.
The development follows a report by the Korea Economic Daily, which stated that Samsung has passed HBM4 qualification tests for both Nvidia and AMD and is preparing to ship to Nvidia starting next month.
Competitive Landscape
SK Hynix announced in October that it had finalized HBM supply agreements with major customers for the coming year. The company is also expanding production capacity, preparing to deploy silicon wafers next month at its new fabrication plant, M15X, in Cheongju, South Korea.
Both South Korean memory giants are scheduled to report fourth-quarter earnings on Thursday, where details of HBM4 orders and production timelines are expected to be disclosed.
Next-Generation AI Chip Pairing
The advanced HBM4 memory is designed to be paired with Nvidia’s forthcoming Vera Rubin AI platform. Nvidia CEO Jensen Huang confirmed earlier this month that the new platform is in “full production” and is slated for launch later this year.
About the Author
