Testing STT-MRAMs
Do We Need Magnets in our Automated Test Equipment?
Sicong Yuan (IMEC-Solliance, TU Delft - Computer Engineering)
Hanzhi Xun (TU Delft - Computer Engineering)
Woojin Kim (IMEC-Solliance)
Siddharth Rao (IMEC-Solliance)
Erik Jan Marinissen (IMEC-Solliance)
Sebastien Couet (IMEC-Solliance)
Moritz Fieback (TU Delft - Computer Engineering)
Mottaqiallah Taouil (TU Delft - Computer Engineering, CognitiveIC)
Said Hamdioui (TU Delft - Computer Engineering, CognitiveIC)
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
The Spin-Transfer Torque Magnetic Random Access Memory (STT-MRAM) is on its way to commercialization. However, the development of high-quality test solutions for STT-MRAMs poses challenges due to the specific working mechanism of the core element of the STT-MRAM bit cells, i.e., the magnetic tunnel junction (MTJ), which involves both a magnetic field and spin-transfer torque. This property can introduce defects unique to MTJs which may escape from test programs that consist solely of functional write and read operations, like march tests. Hence, it is important to develop test solutions that go beyond conventional march tests. This paper explores the effect of applying an external magnetic field (Hext) on the test quality and test time of STT-MRAMs, which could be achieved by integrating one or more magnets in the Automated Test Equipment (ATE) setup. A framework for these so-called Hext-assisted tests is presented and implemented for all known conventional and unique defects. The paper demonstrates that the Hext-assisted tests offer superior coverage and/or lower test time compared to regular functional tests, like march tests. The effectiveness of these tests are validated through silicon measurements.