Creating a design centric ecosystem

To analyze the impact of foundries in an ecosystem, we need to review the semiconductor value chain. In the past, the first four blocks were tightly linked and contributed to the growth of IDMs (Integrated Device Manufacturers) doing the product design (chips) as well as manufacturing (both front-end and back-end). However, recent decades have witnessed the emergence of a business model wherein the product design and manufacturing can be done at an “arm’s length”. This changed business model has given rise to different kind of players like “fab-less’ companies focusing mainly on the product design and system solution on one side, and foundries on the other, focusing on semiconductor manufacturing.

Recent trends show IDMs considering utilization of foundries more as they move towards being “fab light”. Very high capital investment and limited capability to saturate the available capacities during the down turn are among the key drivers behind this trend. Another interesting development is that the foundries are fast emerging as leading players in the development of new technologies, which further provides the reasons for IDM to consider their services.

All this is making foundries critical in the whole semiconductor eco-system. In this context, India being new on the horizon with respect to semiconductor manufacturing, pure “foundries” may be more suited compared to “IDM” setting up their captive chip plants in India.

Possible impact of foundries
Foundries normally come with requirements for much wider skill sets, such as material science to deal with different materials in use in wafers, equipment and control engineering for complex machinery, optical engineering relating to lithography and mask-making, and chemical engineering relating to mediation of chemical interactions during the manufacturing process. Moreover, within the domain of electrical and electronics, there are requirements for high value peripheral support functions intermingled with design skills. These pertain to electrical characterization, modeling, yield and reliability analysis, wafer and product level testing. Consequently, this can bring many more disciplines and associated talent into the fold of semiconductor engineering, and at the same time provide multitude of opportunities to aspiring electrical or electronics engineers.

The presence of foundries can also help accelerate local electronics manufacturing. A matured electronics manufacturing environment does have its direct or indirect impact on the design engineering community starting with system design, which further transmits the influence on the VLSI design. There would be a need for strong system design teams to be able to provide reference designs, service and support to the electronics systems developers and manufacturers. This will bring the much needed end application related knowledge to the very “design centric” Indian design community. This can further help in brining higher levels of effectiveness, and eventually start defining the products most appropriate for the end applications.

The design community
The presence of foundries can also impact the local design community. Today Indian chip designers have limited scope and access to the knowledge of actual semiconductor physics and related challenges. The design starts with the simulatable process parameters (e.g. spice cards) and is handed over at the stage of tape out, i.e. chip’s physical layout. There is limited knowledge both before this process (semiconductor physics, parameter extractions, device modeling, etc.) and after it (testing, manufacturing etc.). With new technologies demanding more and more integration of design and manufacturability, there is a growing need for the designers to be aware of the semiconductor process. Level of yields of the final product can easily become a defining point for a business’ success or failure. Therefore, presence of foundries in the local ecosystem can help in this overall knowledge flow, specifically relating Design for Manufacturability (DFM).

Currently, it is difficult for the educational institutions to access fabs, which further effects the critical possibility of putting research projects on silicon. With the presence of a local foundry, educational institutions can source wafers to test new designs and provide students with projects. This would help upgrade curriculum and align it with latest industry trends. Consequentially, a pool of better-equipped students would enter the semiconductor industry. And finally, this can also facilitate in-depth research on semiconductors in the universities.