Page 40 - Chip Scale Review_July August_2023-digital

PERSPECTIVES IN R&D

A pioneering and successful academic-industry model

in integrated systems packaging at Georgia Tech

By Rao R. Tummala [Georgia Tech]

M any industry-university how universities can perform leading- 127mm substrates. The substrate was

m o d el s h av e b e e n
edge next-generation strategic R&D,
the industry’s first low-temperature co-
developed globally in
standard that is used in all radio-
electronics over the last educate massive numbers of students, fired ceramic (LTCC), now an industry
and co-develop all the technologies
five decades to improve the effectiveness necessary for the next-generation of frequency (RF) applications.
of the academic community, to serve computing, communication, automotive The NSF ERC I started is the flagship
the needs of the local industry, and and consumer systems, and ready them program that funds transformational
to contribute to the competitiveness for commercialization. This paper is technology concepts, such as the
of the applicable country. Academic about how we created an innovative and exploration and development of three
universities are known to perform successful academic/industry model major prog ra ms si mult a neously,
highly-innovative research—but that while developing glass-panel packaging described below.
research hardly ever ends up in products. from concept to commercialization. Explore and demonstrate SoP. This
Industry, therefore, performs its own The result was leading-edge packaging program was in contrast to the use
R&D, independent of academic R&D, for high-performance computing and of SoC packaging, which was being
to address the so called “valley of artificial intelligence (AI), 6G and pursued by the industry at the time.
death” between academic R&D and the beyond communications, and automotive Develop educational programs. A
industry’s manufacturing needs. and consumer electronics. How this was large number of educational programs,
Asia n a nd Eu ropea n cou nt r ies accomplished is the basis of the new including both classroom and hands-
created national industrial institutes model described in this article. on courses, were developed along with
like the Industrial Technology Research curricula and textbooks. A large number
Institute (ITRI) in Taiwan, the Institute Setting-up the Packaging Research of interdisciplinary students participated
of Microelectronics (IME) in Singapore, Center at the BS, MS and PhD levels.
and the Fraunhofer-Gesellschaft in The model we set up is very different Set up and demonstrate a large
Berlin, Germany to address the valley from the other models used by the global industry consortium. It was
of death. Except for imec (Belgium), organizations listed above. The GT PRC essential to gather a large number of
most industrial institutes focus on started as an NSF ERC with a system- researchers, developers, manufacturers
short-term technology development on-package (SoP) vision I defined, and users, all working together to
i n pa r t ne r sh ip w it h i ndu st r y t o which was in contrast to the system-on- co-develop next-generation systems
transfer developed-technologies into chip (SoC) vision widely practiced in technology, ready for manufacturing in
manufacturing. In the U.S., no such the industry. an accelerated mode.
institutes exist as bridges. Companies Georgia Tech is top ranked and the
work directly with universities, mostly largest academic engineering institute Two-part R&D strategy with two
for purposes of training an educated in the U.S. As such, it has a very large R&D teams
workforce. The National Science group of faculty and many, if not all, U.S. universities are viewed as the
Foundation Engineering Research engineering departments, have state- best and most innovative in the world for
Center (NSF ERC) is the best example of-the-art facilities. In semiconductors, leading-edge R&D. But their impact on
of all government-funded programs it has an extensive facility called developing and transferring technologies
in the U.S. that attempts to close the IEN, Institute for Electronics and ready for commercialization has been
gap between academic institutions Nanoelectronics. Within that, it has minimal. In Asian and European
and industry—it starts with the basic many centers including the Packaging countries, this problem has been
research and tries to couple to the Research Center (PRC) created by addressed by universities performing
industry. The Semiconductor Research me when I joined Georgia Tech in research and industrial institutes
Corporation (SRC) was created in the 1993, after 25 years at IBM where I performing technology development in
USA, but its focus is more strategic developed leading-edge packaging, partnership with, and funding from, the
R&D, funded by the industry, and the now called chiplet or 2.5D. This industry, thereby building upon academic
R&D takes place at universities. technology integrated up to 144 small advances.
Georgia Tech Packaging Research chips including logic, memory and Georgia Tech PRC divided R&D into
Center (GT PRC) is an example of capacitor devices, on ver y large, two parts described in the sections below.

38
38 Chip Scale Review July • August • 2023 [ChipScaleReview.com]