ST Microelectronics discusses 28nm R&D
September 1, 2010, Global Technology Conference San Jose, CA—Jean-Marc Chery, Executive Vice President and CTO of ST microelectronics described collaborative R&D model and global multi-fab manufacturing. The trends and challenges driving this model are increases in the logical components within designs and increases in speed of those designs.
Today, companies need to understand where they fit within the value chain, and have to consider multiple fabs for sourcing their production materials. Today’s designs depend on CMOS functionality and characteristics and use multiple libraries to achieve their design goals. Designs regularly use components from the high performance, general purpose, low power, and analog libraries in attempt to make more optimized trade-offs in their designs.
Managing this complexity requires a complete ecosystem. In some cases, general purpose libraries have been displaced with a mixture of high-performance and low-power library elements. Breakdowns are occurring in other areas, requiring greater interactions between technology, designers, and manufacturing. One strategy to address some of these issues is to partner with other companies for R&D.
These partnerships require innovation leadership and the ability to address clusters of competence in each of the companies involved. Collaborative R&D enables companies to share the costs and reduce the risks of new developments. Technology operations of this nature result in a return on investment, in time, and reduced costs in R&D. These two developments also enhance migration of R&D efforts into manufacturing, with reduced risk, time, and knowledge of other issues which could impact the migrations.
Active partnerships and technology developments encourage each company to contribute from their own areas of excellence. Collaboration also lets companies continue to be involved and invest in R&D at levels that are economically responsible. In this way, interactions between technologies and products are encouraged and optimized.
Partnerships in R&D also enable multi-fab manufacturing. If you just work with a foundry, the process is not a differentiator, but if you’re working with a group of other IDM’s, the common processes could provide additional advantages. In this environment, companies must also consider the issues of fab mobility versus the cost of R&D. All of these issues are changing the manufacturing landscape.
In the 65nm through 55nm nodes, the major issue was device matching and migration of manufacturing between fabs was possible. With 32nm processes, the flows are changing into incompatible technologies in the marketplace. To ensure supply security, working with an organization like the common platform group allows for fab synchronization and GDSII compatibility. Collaboration does not preclude competition, but allows for distinct developments at the design level while permitting fab synchronization in manufacturing.
