Extended MTJ TDDB Model, and Improved STT-MRAM Reliability With Reduced Circuit and Process Variabilities

V. B. Naik, J. H. Lim, K. Yamane, Jihye Kwon, Behtash Behin‐Aein, N. L. Chung, S. K, L. Y. Hau, C.J. Chiang, Youda Huang,

2022 IEEE International Reliability Physics Symposium (IRPS) · 2022 · 13 citations · 9 references

Concepts

Abstract

We present a reliable magnetic tunnel junction (MTJ) TDDB model using 40Mb 22FDX <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">®</sup> STT-MRAM at sub-PPM failure rate. This model is based on the precise estimation of voltage across MTJ at bit-cell level derived from compact model and design simulations to cover the product level endurance performance from MTJ diameter, resistance-area product, and temperature effects. We discuss the implications of pre/post MTJ switching, circuit variations and write pulse on MRAM endurance. By using design-process-test co-optimization, we show robust MRAM product reliability to meet >1M cycles with solder reflows and path towards achieving >E12 cycles for cache applications.

References

9