Raj Nair is the Vice President for Technology Development at GlobalFoundries Singapore. The Singapore GlobalFoundries TD group develops embedded NVM, RF and Power Management solutions for a wide range of products and customers. Raj has over 25 years of Operations and R&D experience in the semiconductor and advanced battery industries. He has held executive positions at Johnson Controls, Maxim, Motorola and On Semiconductor. At Johnson Controls he was the Vice President for Advanced Manufacturing and Quality and launched the world’s largest automotive Li ion battery plant in Holland, MI. He is a pioneer in the development of Smartpower integrated circuit development. Raj has a B.Tech degree from IIT Bombay; a Master of Science (EE) from NTU and a Master of Science (ChE) from Syracuse U.

Abstract

Advanced mobile applications are the predominant driver for semiconductor technology innovations both at the leading edge (“More Moore”) and at the mature nodes (“More than Moore”). The short life-cycles of mobile products and the need for SoC level differentiation puts significant challenges on technology architecture and time-to-volume. We foresee a move towards much closer collaboration, in fact a virtual IDM-like model, on both technical and business levels. With daunting technical challenges like 3D stacking, 450mm fabs, new transistor architectures, multi-patterning, extreme ultraviolet (EUV) lithography, collaboration – early, often and deep - is really the only practical approach given the cost and complexities involved. Early eco-system enablement with accurate models and silicon proven IP ensure first pass design success, which is an important step for faster Time-to-volume (TTV). Use of Advanced Process Control (APC) techniques, Design for Manufacturing processes that go beyond the traditional OPC (Optical Proximity Correction) and process corners analysis, and systems that ensure learning from one product are transferred to the next are all critical to a rapid yield learning curve. We refer to this close collaborative model as Foundry 2.0.