Extending DTCO For Today’s Competitive IC Landscape

As semiconductor components continue to shrink, the challenges associated with design-for-manufacturing (DFM) and design-technology co-optimization (DTCO) increase. The complexity of the IC design and manufacturing process demands an extension of traditional DFM and DTCO techniques to overcome the systematic failures tied to complex design-process interactions.

Designers need to accelerate development, manufacturers needs to ramp yield quickly and maximize the efficiency of the fab, and all parts of the supply chain need to be in sync. For the semiconductor fab, this means implementing a digital twin to reduce iterations between design and production, virtually prototype and develop new processes, and quickly characterize and calibrate every piece of equipment in the fab.

The IC design-to-manufacturing flow has well-defined modules such as physical design, mask synthesis, mask writing, fab process and inspection and test. Each module includes industry-standard verification processes, such as physical verification, optical process correction and mask proximity correction (OPC/MPC) verification, metrology inspection and physical failure analysis (PFA). The communication between these modules is one-directional handoff, as shown in figure 1.

Fig. 1: Design-to-manufacturing and final die level diagnosis and physical failure analysis cycle.

The entire process from design to tested chip usually takes somewhere between 6-18 months from start to finish, which is an exceptionally long time to wait for the feedback loop to close. The costs associated with this on-way feedback are high and can negatively impact the entire project.

Traditional DTCO
DTCO is critical for yield breakthrough and product ramp-up during the life cycle of a new technology node. It involves the simultaneous optimization of the chip design and manufacturing process, considering the interdependence between the two. It uses iterative design and manufacturing simulations to identify optimal design and manufacturing parameters.

Later in the process node’s lifecycle, such co-optimization is done by traditional techniques like design-for manufacturability (DFM) and litho-friendly design (LFD). But with the shrinking manufacturing technology and growing design complexity comes greater challenges that existing DFM and DTCO can’t overcome.

Traditional DTCO, DFM and LFD approaches all have proven value, but many designs need more. The truth is that ever more systematic defects escape traditional detection methodology and show up during the yield ramp and final high-volume manufacturing (HVM). To improve the efficiency and effectiveness of the DFM and DTCO processes for complex, advanced node designs, the industry needs more than piecemeal solutions or band-aids.

Building a digital twin for semiconductor fabs
The goal of smart manufacturing using a digital twin approach is accelerate the transfer between products and production, and also back from production to product. With a digital twin, fabs can characterize every piece of equipment in the fab. Making use of all the data a fab generates will narrow the iterative efforts between design and production and allow for quick virtual prototyping and development of new manufacturing processes. Figure 1 illustrates the concept of a digital twin implementation, in which the digital world accelerates the design, optimization and realization of both the product and the production of semiconductors. A digital twin implementation allows for realistic simulations and validation of product, machines, lines, and complete fabs and provides the foundation for flexible and efficient manufacturing.

Extended design-technology co-optimization
Because DTCO is so critical to the definition of a process node, Calibre group at Siemens EDA started researching ways to extend DTCO further, to the entire design-to-manufacturing process by creating a robust information channel. This infrastructure helps get the massive amounts of data where it needs to go to be useful. Our improved and extended DTCO solutions is based on a systematic information exchange and analytics platform that spans various design-to-manufacturing modules with complete lifecycle support. It offers methodologies and infrastructure needed to feed pre-silicon design data and intelligence forward into the manufacturing process, and feed manufacturing information back, post-silicon, to inform the design process. These technologies are available in the Fab Solutions product family.

Fig. 2: Extended DTCO flow, which uses ML to give easier access to patterns, locations, properties, distribution, etc. across modules.

The Fab Solutions products address related issues, such as managing the massive amounts of data generated through the flow and how to bring design content in to process optimization.

Conclusion
The semiconductor industry will benefit from extending DTCO from technology development to HVM. Implementing far-reaching design and process co-optimization has already shown promising results in several applications, suggesting that it may lead to a new era of extended DTCO. Calibre Fab Solutions are leading the way in the use of ML technology to create a digital twin approach that extended DTCO and digitizes the entire fab production process. The improvements in design and process co-optimization using ML will lead to remarkable improvements in the quality, speed, and efficiency of semiconductor manufacturing.