ERC Researchers Demonstrate Memory Element with Ultra-Fast Speed and Superb Energy Efficiency

Achievement date: 
2014
Outcome/accomplishment: 

Researchers at the NSF-funded Nanosystems Engineering Research Center (ERC) for Translational Applications of Nanoscale Multiferroic Systems (TANMS), which is led by the University of California Los Angeles (UCLA), have demonstrated a working memory element that has ultra-high performance in terms of energy efficiency and ultra-fast access speeds.

Impact/benefits: 

This achievement represents a major step toward the ERC's goal to revolutionize the development of consumer electronics by engineering materials that optimize energy efficiency, size, and power output on the small scale. Specifically, with regard to memory devices, the researchers demonstrated an element that is 10x more energy efficient than state of the art (plus a roadmap to improve further to 1,000x more efficient with a combination of materials and mechanical effects) and has write times better than one billionth of a second (1 ns).

Explanation/Background: 

Conventional memory devices generate magnetic fields as currents pass through wires, but this approach becomes very limited as sizes shrink to an extremely small scale (i.e., nanoscale). TANMS seeks to overcome the limitations of the current-through-wire approach by using materials that react in a manner analogous to putting a voltage-controlled on-off switch on a permanent magnet. The ERC aims to demonstrate this new approach by using miniature memory devices in an upcoming testbed system.

Representative switching measurements have shown that the single bit of memory tested meets all requirements of a functional memory bit proposed for the future TANMS memory testbed system. To date, extensive single-bit testing has been accomplished, including tests on write energy as well as tradeoff studies among key device performance parameters, such as write voltage, speed, and energy. The measured experimental data are in good agreement with modeling results, which in turn are being used for improving circuit designs required for the upcoming testbed. Research during the last year has also yielded significant improvements in material stack designs, leading  to better thermal stability (retention time).

In practical memory applications a wide range of operating temperatures need to be accommodated by the memory chip. This system-level issue needs to be addressed early because it is an important consideration defining performance and design specifications. For this reason, the researchers performed a detailed measurement and analysis of aspects of temperature dependence in magnetic memory devices. Discoveries during this testing provide some fundamentally new technical information about responses of the devices.