SK Hynix announced on October 27 that it showcased its strategy for next-generation NAND storage products at the 2025 Open Compute Project (OCP) Global Summit held in San Jose, California, from October 13 to 16 (local time).
The company explained, “With the rapid growth of the AI inference market, demands for NAND storage products capable of processing large amounts of data quickly and efficiently are expanding significantly. To meet this demand, SK Hynix has established the ‘AIN (AI-NAND) Family’ lineup, designed as optimized solutions for the AI era.”
During the executive session held on the second day of the event, Vice President Kim Cheon-seong, responsible for eSSD product development, introduced the AIN Family.
The AIN Family comprises NAND solutions optimized in three key aspects: performance, bandwidth, and density. It aims to enhance data processing speeds and maximize storage capacity.
AIN P (Performance) is tailored for large-scale AI inference environments, efficiently handling vast data input/output. It aims to minimize bottlenecks between AI processing and storage, significantly improving speed and energy efficiency. SK Hynix is currently designing new structures for NAND and controllers, with samples expected to be released at the end of 2026.
AIN D (Density) is a high-capacity solution focused on low power consumption and low cost, suitable for storing large volumes of data. It seeks to surpass existing QLC-based TB (terabyte) SSDs by creating a medium-tier storage that offers capacities up to PB (petabytes) while providing SSD speed and HDD affordability.
(Note: QLC—Quadruple Level Cell—refers to NAND flash technology storing four bits per cell, with other standards like SLC, MLC, TLC, and PLC representing fewer or more bits per cell.)
AIN B (Bandwidth) is a solution that stacks NAND layers to expand bandwidth, employing the ‘HBF™ (High Bandwidth Flash)’ technology, which stacks NAND similar to DRAM-based HBM (High Bandwidth Memory).
SK Hynix, with its leading expertise in developing and manufacturing high-performance HBM, has long been addressing the memory capacity issues arising from expanding AI inference and large-scale LLMs (Large Language Models). Its core strategy involves integrating high-capacity, low-cost NAND with HBM stacking structures. The company is exploring various applications by combining AIN B with HBM to address capacity challenges.
To expand the AIN B ecosystem, SK Hynix, in partnership with U.S. company SanDisk through a memorandum of understanding (MOU) on HBF standardization signed in August, hosted an event called ‘HBF Night’ at The Tech Interactive near the OCP venue on the evening of October 14, inviting global big-tech executives.
The event, attended by domestic and international professors and industry leading engineers, featured panel discussions on accelerating NAND storage innovation and industry-wide collaboration.
Hynix’s Chief Development Officer, Ahn Hyun, stated, “The OCP Summit and HBF Night showcased SK Hynix’s current and future position as a ‘Global AI Memory Solution Provider’ amid rapidly evolving AI-centric markets. We will continue to work closely with customers and partners in next-generation NAND storage development to become a key player in the AI memory industry.”
