Performance of spin orbit torque-driven electronic synapse function in perpendicularly magnetized multilayers
- Title
- Performance of spin orbit torque-driven electronic synapse function in perpendicularly magnetized multilayers
- Other Titles
- 수직 자기 다층 구조에서 spin orbit torque를 이용한 시냅스 소자 특성 확인
- Author
- 신정훈
- Advisor(s)
- 홍진표
- Issue Date
- 2019-02
- Publisher
- 한양대학교
- Degree
- Master
- Abstract
- Processing machine faster than human’s brain has been leading a society in computing for decades, and the Von Neumann architecture has become the clear standard for such a machine. However, the unavoidable comparisons of this architecture to the human brain significant differences in the organizational structure, low power, and recognition between the two. This leads to a natural question regarding the feasibility of creating alternative architectures based on neurological models that compare favorably to a biological brain. Hardware development is in progress to replace the Von Neumann structure. In particular, magnetic devices utilizing spin orbit torque through magnetic manipulation by in-plane current injection have become on candidate of alternatives to conventional devices due to lower power consumption and faster magnetization switching. Therefore, we report performance of nanoscale electronic synapses which are based on spin orbit torque (SOT) phenomena observed from the modified magnetic tunnel junction (MTJ) configuration. The SOT-driven domain wall motion (DWM) was observed by means of the application of in-plane current pulse injection, confirmed by the Magneto-Optic Kerr Effect (MOKE) analyses. A periodic repetition of writing and reading pulses indicated a variation in resistance, reflecting a typical synapses function. All of this process was confirmed in the free layer and confirmed in the full structure.
- URI
- https://repository.hanyang.ac.kr/handle/20.500.11754/100374http://hanyang.dcollection.net/common/orgView/200000434664
- Appears in Collections:
- GRADUATE SCHOOL[S](대학원) > PHYSICS(물리학과) > Theses (Master)
- Files in This Item:
There are no files associated with this item.
- Export
- RIS (EndNote)
- XLS (Excel)
- XML