Samsung has just announced a new non-volatile memory chip technology that could be a major breakthrough in the world of semiconductors. This technology is called Ferroelectric Field-Effect Transistor or FeFET wich is claimed to be able to increase storage capacity without increasing power consumption. Will it replace NAND in the future?
Samsung Develops FeFET Chip with 96% Power Reduction, Poised for AI Applications
Samsung researchers have developed a new Ferroelectric Field-Effect Transistor (FeFET)-based memory system utilizing a combination of hafnia and zirconium, potentially revolutionizing data storage with significantly reduced power consumption.This technology promises to address a key limitation of customary NAND flash memory – its high power demands.
The core issue with conventional NAND chips lies in their structure. Multiple memory cells are interconnected, requiring power to be supplied to all cells even when only one is being accessed. This leads to increased power consumption, especially as chip capacity grows.https://www.jagatreview.com/samsung-develops-fefet-chip-with-96-power-reduction-poised-for-ai-applications/
Samsung’s FeFET solution overcomes this by employing a ferroelectric material that retains its electric charge even without a power supply. This inherent property results in a more efficient and stable chip capable of storing up to five bits of data per cell while drastically reducing energy usage.
Testing indicates the new FeFET technology can reduce power consumption by up to 96% compared to standard NAND chips. https://biz.chosun.com/en/en-science/2025/11/27/YCNPCOAZQBETVLOAZR2ABXGF6E/ Furthermore, FeFETs demonstrate stability in small, three-dimensional designs at a scale of 25 nanometers, paving the way for denser and faster chips.
Samsung identifies FeFET as a next-generation non-volatile memory technology offering both high capacity and extraordinary power efficiency.While currently in the research and advancement phase, the technology holds critically important promise, especially for applications demanding large-capacity storage, such as the rapidly growing artificial intelligence (AI) industry.
Worth a look