Micron previewed its G11 NAND flash at its 2026 Technology Webinar, with capacity ramp expected in H2 2027. The company appears to have skipped the G10 generation.
G11 NAND stacks over 400 layers, uses wafer-to-wafer bonding, and delivers roughly 2x the areal storage density of current 276-layer G9 NAND, while enabling up to 1.5x I/O rates and reducing data transfer energy by over 25%. After G11, Micron plans to introduce high-mobility materials into NAND to enable ultra-dense cell architectures and further strengthen energy efficiency.
Micron also introduced two AI-grade data center NAND designs: an SLC-based GPU-direct IOPS-optimized product and a bandwidth-optimized product similar to the HBF definition. The company also showcased its full range of AI-optimized storage solutions.
ICgoodFind Summary: Micron skipped G10 to push G11 forward, using wafer-to-wafer bonding to exceed 400 layers—doubling density while cutting power. AI data center demand for NAND IOPS and bandwidth is splitting into dedicated product forms, shifting memory makers' competition from capacity to scenario fit.
