SK hynix starts supplying high-performance NAND for mobiles

Seoul [South Korea], September 11 : SK hynix Inc. said on Thursday that it began supplying its ZUFS (Zoned UFS) 4.1 solution, the world's first high-performance NAND for mobile devices, to customers for boosting AI and large-scale data processing, as per a report by Pulse, the English service of Maeil Business Newspaper Korea.

ZUFS applies zoned storage technology, which stores data by usage and characteristics, to high-speed flash memory UFS (Universal Flash Storage). Working closely with clients, the company developed version 4.1, which features significantly improved data transfer speeds and power efficiency. Mass production began in July 2025 following certification the month before.

The product is already being integrated into the latest smartphones, demonstrating the company's technological strength in the global market. The company said that it expects the new solution to support powerful on-device artificial intelligence functions and deliver innovative user experiences.

"When installed in smartphones, ZUFS 4.1 improves operating system speed and enhances data management efficiency. Over extended use, read performance degradation is reduced by more than fourfold, cutting app launch times by 45 per cent compared with conventional UFS. AI app execution time is also reduced by 47 per cent, while error handling has been substantially strengthened over the previous version, boosting system reliability and recovery," the report said.

SK hynix explained that these improvements make the new product well suited to today's mobile environment, where on-device AI and large-scale data processing are essential.

The company also noted that ZUFS 4.1 is the first product co-developed and mass-produced through collaboration to optimize the Android operating system and storage devices. SK hynix said it will continue to supply NAND solutions in line with customer needs while strengthening partnerships with global firms to build differentiated competitiveness in the AI memory sector.

Disclaimer: This post has been auto-published from an agency feed without any modifications to the text and has not been reviewed by an editor