Ring-shaped Racetrack memory based on spin orbit torque driven chiral domain wall motions

Area of Science:

• Spintronics
• Materials Science
• Computer Engineering

Background:

• Racetrack memory (RM) offers potential for low-power, high-density, high-speed data storage.
• Conventional stripe-shaped RM faces data overflow and reliability issues due to bidirectional domain wall (DW) shifting.
• Geometric constraints in stripe RM increase circuit complexity and programming challenges.

Purpose of the Study:

• To propose and investigate a novel ring-shaped racetrack memory (RM) architecture.
• To address the data overflow and reliability limitations of conventional stripe-shaped RM.
• To explore the application of spin orbit torque (SOT) driven chiral DW motions in RM.

Main Methods:

• Utilized micromagnetic simulations to validate the functionality of ring-shaped RM.
• Investigated spin orbit torque (SOT) driven chiral domain wall (DW) motions.
• Modeled current flow through a heavy metal layer for DW shifting.

Main Results:

• Ring-shaped RM successfully demonstrates functionality and performance advantages over stripe designs.
• The 'end to end' circulation of storage data is achieved by current flowing through the heavy metal.
• All data remains within the device during shifting, preventing data loss.

Conclusions:

• Ring-shaped RM based on SOT-driven chiral DW motion effectively resolves data overflow issues.
• The proposed architecture enhances data integrity and reliability in memory devices.
• This work presents a promising pathway for practical memory and logic applications of RM.