Library Subscription: Guest
Begell Digital Portal Begell Digital Library eBooks Journals References & Proceedings Research Collections
Computational Thermal Sciences: An International Journal
ESCI SJR: 0.244 SNIP: 0.434 CiteScore™: 0.7

ISSN Print: 1940-2503
ISSN Online: 1940-2554

Computational Thermal Sciences: An International Journal

DOI: 10.1615/ComputThermalScien.2013007643
pages 441-458


Avram Bar-Cohen
Laboratory of the Thermal Management of Electronics, Department of Mechanical Engineering, University of Minnesota, Minneapolis, MN 55455; and Defense Advanced Research Projects Agency (DARPA), Microsystems Technology Office, University of Maryland, College Park, MD
Ankur Srivastava
Electrical and Computer Engineering, University of Maryland, College Park, Maryland, USA
Bing Shi
Electrical and Computer Engineering, University of Maryland, College Park, Maryland, USA


This paper focuses on the thermal challenges in three-dimensional (3D) chip stacking, which is poised to become the next packaging paradigm in the electronic industry. While this paradigm provides significant improvements in device density, interconnect delays, and system integration, it is expected to lead to higher heat densities along with decreased access of chip hot spots to air-cooled heat sinks. Excessive chip temperatures are a well-known culprit leading to performance and reliability loss as well as higher leakage power. This paper investigates the advantages and challenges imposed by interlayer microfluidic cooling to address the thermal needs of 3D integrated circuits (ICs). We present a brief review of the emerging 3D form factors and application of microfluidic cooling followed by several case studies that quantify the potential improvement obtained by co-design of the thermal and electrical aspects of the 3D ICs. We underscore the need for better unification of the thermal and fluidic aspects of the system into an integrated co-design environment that enables designers to estimate the impact of the cooling solutions on the electrical aspects and vice versa. Such a co-design approach would be imperative for unlocking the high performance and energy efficiency of 3D ICs.