The economic downturn has been tough on the entire semiconductor industry. Severe cost-cutting measures have slashed staff in design and CAD organizations. Meanwhile, semiconductor companies still face competitive pressure to deliver higher-performance, lower-power products. To meet those expectations, companies are investing in support for nanometer process technologies, but tighter budgets force hard choices regarding which core competencies and infrastructure projects are affordable.

The growing number of casualty reports suggests that more-sophisticated design flows for nanometer effects at 130 nm and below are critical to organizational success. Electrical effects related to process scaling include increases in coupling noise, leakage current and on-chip process variation. To counter them, design teams have adopted techniques, such as voltage islands and well biasing, that further complicate the design process.

These challenges require advanced design tools; however, the analysis accuracy of these tools depends on the availability and quality of new models. Therefore, leading semiconductor companies find that advanced characterization and modeling are required for successful nanometer design.

Design teams recognize this fact, but often the library teams serving them rely on internally developed characterization systems. Built to support static timing analysis using simple lookup table models and later extended for switching power, these legacy systems were well-suited to the less-stringent timing and power characteristics of submicron processes. With the advent of nanometer design, however, not only is completely new data for noise and leakage required, but the models themselves are becoming more complex. For instance, equation-based models are replacing lookup tables to support on-chip variation, as well as multivoltage and multithreshold design.

To address the new characterization and modeling issues, library teams are faced with a retooling process. The significant time and resource requirements involved in upgrading existing characterization infrastructure force new make-vs.-buy decisions.

With strong competitive pressures on product development, prudent companies are acting now to acquire a commercial solution to replace home-grown characterization systems.