In the big data era, the use of consumer electronic products in all aspects of our life is manifested by the ubiquitous presence of mobile devices. The continuing demand for smaller size, more functionality, lower power consumption, and reduced cost is challenging. A paradigm change, from 2D IC to 3D IC, is occurring in microelectronic industry. The stacking of thin chips makes Joule heating to be the most serious yield and reliability issue. However, to remove heat a temperature gradient must exist, but a difference of 1°C across a μ-bump of 10 µm in diameter has a temperature gradient of 1000 °C/cm, which can cause thermomigration. Furthermore, system level electromigration occurs in weak links, and fracture occurs in intermetallic μ-bumps. In this talk, how to use artificial intelligence (AI) for faster reliability study will be discussed.
[1]
W. Carter,et al.
Irreversible Thermodynamics: Coupled Forces and Fluxes
,
2005
.
[2]
Menglu Li,et al.
Effect of Joule heating and current crowding on electromigration in mobile technology
,
2017
.
[3]
Harold Barnard,et al.
Joule heating induced thermomigration failure in un-powered microbumps due to thermal crosstalk in 2.5D IC technology
,
2016
.
[4]
Ilya Prigogine,et al.
Introduction to Thermodynamics of Irreversible Processes
,
1967
.
[5]
K. Tu,et al.
Synergistic effect of electromigration and Joule heating on system level weak-link failure in 2.5D integrated circuits
,
2015
.