Micron plans double-density memory chips for late 2027
The company confirmed it will use wafer bonding to reach 42 Gb/mm2, a density that exceeds current competitors.
By Malik Salazar ·
Micron announced this week that its next generation of 3D NAND flash, designated G11, will begin high-volume production in the second half of 2027. The company stated that the new chips will offer approximately twice the storage density of its current generation while improving speed and power efficiency. This rollout marks a major shift in the company's manufacturing strategy as it seeks to maintain its position in the memory market.
The technical specifications show a clear increase in performance metrics. Micron projects that G11 will reach a density of 42 Gb/mm2 in triple-level-cell form, up from the current 21 Gb/mm2. Input/output speeds are expected to rise by 50%, moving from 3,600 MT/s to 5,400 MT/s. The energy required to transfer each bit should drop by more than 25 percent. These figures place Micron ahead of rivals Kioxia and Samsung, whose next-generation products are expected to have lower density.
Scott DeBoer, president and chief technology and products officer at Micron, confirmed the design changes at an industry event. He described G11 as the world's most advanced NAND and verified that it relies on wafer-on-wafer bonding. DeBoer noted that the company has long planned to adopt this technology when it became cost-effective. He stated, 'This is the right timing for that technology.'
The move to wafer bonding allows Micron to construct the memory array and the peripheral circuitry on separate wafers. Each wafer is manufactured using a process optimized for its specific function before the two are bonded together. This approach departs from the company's previous CMOS-under-array designs, granting engineers greater flexibility to optimize both the memory and logic components independently.
Micron confirmed that G11 will feature more than 400 layers, though it did not disclose the exact layer count or cell pitch details. The company indicated that the density gains will come from a combination of additional layers and smaller cell dimensions. More technical specifics are expected to be released in 2027, closer to the start of mass production.
The new chips are intended for high-capacity solid-state drives used in client and data center applications, as well as storage devices for artificial intelligence workloads. These AI systems require higher bandwidth and better energy efficiency. Micron stated it will continue scaling in all three dimensions beyond G11, using new materials and denser cell architectures to improve performance.
DeBoer mentioned that the company has a strong roadmap for future generations but did not provide specific dates or density targets for products after G11. He emphasized a focus on new materials and continued scaling. Micron has not yet shared architectural details for any generation following G11.
