Taking a novel approach
Researchers at the Korea Advanced Institute of Science and Technology (KAIST) are making strides towards a significant memory breakthrough with the development of a new memory device that merges the characteristics of DRAM and NAND flash memory. This innovative technology has the potential to offer a more cost-effective production method and greatly reduce power consumption, addressing key challenges in the current memory landscape.
Professor Shinhyun Choi, leading the team at KAIST’s School of Electrical Engineering, is at the forefront of this groundbreaking project. The research aims to create memory solutions that are not only more efficient but also have the potential to replace existing memory technologies or be integrated into future AI hardware for neuromorphic computing.
The newly designed memory device utilizes phase change memory technology with ultra-low power consumption, offering a viable alternative to both DRAM and NAND flash memory. By combining the speed of DRAM with the non-volatility of NAND flash memory, this innovation represents a significant leap forward in memory development.
Previous iterations of phase change memory technology suffered from high power consumption, limiting their practicality in real-world applications. However, Professor Choi’s team has managed to overcome this hurdle by devising a method to electrically form phase change materials in a highly compact space, resulting in an ultra-low-power memory device. This breakthrough technology consumes 15 times less power than earlier phase change memory models, marking a major advancement in cost-effective and energy-efficient memory solutions.
Professor Choi emphasized the significance of this development, stating that it offers a fresh approach to addressing longstanding challenges in memory device production. He envisions this research laying the groundwork for future advancements in electronic engineering, particularly in the realms of high-density three-dimensional vertical memory and neuromorphic computing systems.
KAIST’s focus on neuromorphic computing extends beyond this project, as evidenced by the recent unveiling of an AI chip that rivals the speed of Nvidia’s A100 GPU while boasting a smaller form factor and significantly lower power consumption. These initiatives underscore the university’s commitment to driving innovation in the field of memory and computing technology.