Samsung Electronics is making an aggressive push into next-generation high-bandwidth memory, pivoting its entire AI storage chip strategy toward high-value-added products amid intensifying competition in the sector.

According to reports from Seoul Economic Daily, Samsung plans to increase outsourced cleaning volumes for glass carriers — an essential component in the HBM manufacturing process — from 20,000 wafers per month this year to 50,000 per month next year, a 2.5-fold jump. Industry insiders say this points to HBM4 and HBM4E actual production volumes at least doubling next year compared to this year. At the same time, Samsung's overall monthly HBM wafer input is expected to expand from roughly 180,000 wafers this year to approximately 250,000 wafers next year, representing growth of nearly 40%.

Behind this production expansion lies a strategic concentration on advanced product generations. As HBM4E ramps toward mass production, the share of HBM4 series in total HBM shipments is expected to surge from around 40% this year to approximately 80% by next year. This structural shift signals that Samsung will redirect more production capacity toward higher-margin products.

Surge in Glass Carrier Demand Reflects Capacity Expansion Logic

Glass carriers are a critical consumable in the HBM production process. During the grinding, thinning, and drilling of DRAM wafers, the glass carrier is attached to the backside of the wafer to prevent warping or cracking. Since HBM requires stacking multiple DRAM layers within a limited thickness, the importance of wafer thinning technology and warpage control processes rises as stacking layers increase. Samsung's key expansion targets — HBM4 and HBM4E — both center on products with 12 or more high-stack layers, imposing even more stringent process requirements on glass carriers.

According to Seoul Economic Daily, Samsung's monthly glass carrier demand has grown from 10,000 wafers last year to 20,000 wafers this year, and is expected to further increase to 50,000 wafers next year. Industry analysts note that even accounting for the fact that glass carriers can be reused after cleaning, and that actual consumption varies depending on process input methods and yield rates, the 2.5-fold demand surge still strongly points to expectations of HBM4 series production more than doubling.

Product Structure Accelerating Transition to Advanced Generations

On the mass production front, Samsung launched HBM4 volume shipments in February this year. The product utilizes 10-nanometer-class sixth-generation (1c) DRAM and a base die manufactured using a 4-nanometer process. In May this year, Samsung also provided samples of the HBM4E 12-layer version to customers including Nvidia.

Industry projections indicate that as HBM4E enters full-scale mass production next year, the HBM4 series' share of Samsung's total HBM shipments will climb sharply from approximately 40% this year to around 80%. One industry insider commented: "While expanding HBM production, Samsung appears to be positioning the high-value-added HBM4 at the core of its strategy." This shift in product structure will directly enhance the overall profitability of Samsung's HBM business and strengthen its competitive position in the AI accelerator memory market.

High-Stacking Process Demands Drive Supply Chain Momentum

Samsung's latest expansion round is creating significant pull-through effects on its upstream supply chain. Glass carrier outsourced cleaning volumes are set to expand from 10,000 wafers per month to 50,000 per month within two years — a fivefold increase — with related suppliers' order volumes scaling up correspondingly.

From an industry trend perspective, the continuous increase in HBM stacking layers — evolving from the early 8-layer and 12-layer configurations toward even higher counts — is making wafer thinning and support processes increasingly pronounced bottlenecks in the overall manufacturing flow. The rapid growth in glass carrier demand is both a direct reflection of Samsung's capacity expansion and an indication of the rising material consumption intensity driven by the growing complexity of high-stacking HBM processes.