Project description
Low-power non-volatile memory based on spin-orbit torque switching
Most parts of modern computers are made with volatile devices such as transistors, which lose information when powered off. Requiring no power to maintain stored data, magnetoresistive random-access memory (MRAM) is an emerging class of non-volatile memory that opens the door to a new paradigm based on normally-off and instant-on operation. What is more, this type of memory reduces the gigantic energy losses associated with data storage and leakage power consumption. The EU-funded SOTMEM project aims to demonstrate advanced MRAM elements whose state will be controlled by spin-orbit torque magnetisation switching. Spin-orbit torque MRAM provides a route around the limited speed as well as reliability and degradation issues that spin-transfer torque MRAM faces. The project is set to accelerate the transition towards low-power and secure microprocessors with non-volatile memory.
Objective
SOTMEM addresses the growing need for scalable ultrafast non-volatile memories (NVM) to improve the reliability of logic circuitry and to reduce the ever increasing power consumption and energy loss in microprocessors. Major technology actors and end-user companies are developing magnetic RAM (MRAM) as it is recognized as one of the most promising emerging NVMs. However, mainstream MRAM, relying on spin transfer torque, suffers from limited speed, and reliability and degradation issues. These three obstacles, which impede the widespread implementation of MRAM, can be mitigated using a novel device architecture based on spin-orbit torque (SOT). But even SOT-MRAM brings challenges in the form of too large writing current density and power dissipation . SOTMEM will validate SOT switches for SOT-MRAM using topological insulators embedded in a novel material stack that we have recently developed for optimal writing efficiency. Writing at ultralow power is therefore the key technical objective of this ERC Proof of Concept project. A successful outcome would represent a major breakthrough for SOT-MRAM commercialization. Therefore SOTMEM stands to have enormous impact in the transition towards low-power, power-fail protected microprocessors with non-volatile memories. Special efforts will be devoted to market analysis and patentability studies, as well as targeted industry engagement to allow the design of a nuanced business model aiming to deliver SOTMEM technologies and knowhow to potential end users.
Fields of science
Not validated
Not validated
Programme(s)
Funding Scheme
ERC-POC-LS - ERC Proof of Concept Lump Sum PilotHost institution
08193 Cerdanyola Del Valles
Spain