Statistical Methods for Semiconductor Chip Design

The Open Modeling Coalition (OMC) has released a document titled “Statistical Methods for Semiconductor Chip Design.” The document was developed by the OMC’s Statistical Working Group, and is freely available to the industry on the Si2 web site link as noted below. The OMC is a project under the auspices of the Silicon Integration Initiative (Si2), and the OMC member list is provided below. This significant public resource is intended to help guide industry in the understanding and application of statistical techniques in chip design flows, engaging broad collaborative alignment of companies spanning semiconductor, EDA, and IP industry segments.

This document concerns itself with statistical techniques needed to ensure the progression of Moore’s law. Lithographic and other process techniques are becoming more and more limited in what can be accomplished from one generation to another. To ensure that scaling can continue to provide the semiconductor industry and its customers the required improvements in performance, better statistically based tools and techniques need to be applied so that design specific requirements can be optimized around technology specific effects.

The document outlines “best practice” techniques that are being adopted by much of the industry regarding statistical design and how such methods fit into a typical design flow. Topics covered include:

1. Need for Statistical Techniques in a Design Flow
1.1 Definitions and Acronyms
1.2 Case Study, Technology Trends & Limitations in Lithography
1.3 Case Study, Statistical STA vs. Conventional STA
1.4 Case Study, Chip Timing and Optimization Using Statistical Analysis Methods
2. Intended Usage of Statistical Techniques in a Design Flow
2.1 Design Flow Description, Concept to Reality
2.2 Initial Intent of Specification
3. Specification Details
3.1 Physical Variation Effects & Their Approximations
3.2 Mathematical Modeling for SSTA
3.3 Timing Propagation

This document is intended to advance the understanding in the industry of statistically based methods and to encourage and support their implementation and use. Following is a link to the document: “Statistical Methods for Semiconductor Chip Design”: www.si2.org/?page=1003.

To hear more about the OMC and other Si2 activities, everyone is invited to attend the Si2 Member/Guest meeting at DesignCon on February 2, for more details: www.si2.org/?page=11.

About the Open Modeling Coalition (OMC)
The OMC technical objectives are to define a consistent modeling and characterization environment in support of both static and dynamic library representations for improved integration and adoption of advanced library features and capabilities, such as statistical timing. The system will support delay modeling for library cells, macro-blocks and IP blocks, and provide increased accuracy to silicon for 90nm and 65nm technologies, while being extensible to future technology nodes. Member companies are: Advanced Micro Devices (NYSE: AMD), Altos Design Automation, ARM (Nasdaq: ARMHY), Cadence Design Systems (Nasdaq: CDNS), Fenix Design Automation, IBM (NYSE: IBM), IMEC, Nangate A/S, NXP Semiconductors, Renesas Technology Corp.

About Si2
Si2 is an organization of industry-leading semiconductor, systems, EDA, and manufacturing companies focused on improving the way integrated circuits are designed and manufactured in order to speed time to market, reduce costs, and meet the challenges of sub-micron design. Si2 is uniquely positioned to enable collaboration through a strong implementation focus driven by its member companies. Si2 focuses on developing practical technology solutions to industry challenges. Si2 represents nearly 100 companies involved in all parts of the silicon supply chain throughout the world. Web site: www.si2.org.