SK hynix Inc., announced today that it has developed the Zoned UFS, or ZUFS1 4.0, a mobile NAND solution product for on-device2 AI applications.This new product is set to redefine the capabilities of mobile devices like smartphones, marking a significant leap in the industry.
The ZUFS 4.0 is engineered to optimize on-device AI performance, a feature that SK hynix claims sets it apart as the best in its class. The innovation is expected to to help expand its AI memory leadership to the NAND space, extending its success in the high-performance DRAM represented by HBM.
The ZUFS technology operates by categorizing and storing data from smartphones into different zones based on their characteristics. This method is a departure from traditional Universal Flash Storage (UFS) systems, which do not differentiate data.
By grouping and storing data with similar purposes and access frequencies in separate zones, ZUFS enhances the operating system speed and storage management efficiency of smartphones.
The ZUFS also shortens the time required to run an application from a smartphone in long hours use by 45%, compared with a conventional UFS. With the issue of degradation of read and write performance improved by more than four times, the lifetime of the product also increased by 40%.
The development of the technology dates back to 2019, before the arrival of the AI boom, when SK hynix started collaboration with a global platform service company on expectations for an increased demand for high-performance NAND solutions.
Ahn Hyun, Head of the N-S Committee at SK hynix, said that customers are requiring better memory options as big tech companies focus on development of the on-device products adopting their own generative AI applications. “SK hynix will continue to work toward strengthening its leadership as the global top AI memory provider by supplying high-performance NAND solutions that meet such higher requirements at a right time, while building up stronger partnership with leading ICT companies.”
Also Read: LTIMindtree and IBM Collaborate on watsonx Center of Excellence for Generative AI
