Archive for May, 2009

The Design Automation Standards Committee (DASC) is up to all sorts of good.  Its ongoing collaboration with the IEEE Standards Association’s Corporate Advisory Group (CAG) and plans to reach out to other standards-setting organizations in the design automation industry are indicative of its vitality as a sponsor of IEEE standards.

The DASC has been in existence for decades, and it’s important that it continues to evolve with changing times. As a “steering committee” of sorts for the majority of IEEE design automation standards, it approves and guides standards through both entity- and individual-based processes.

DASC has a goal of mutual cooperation with CAG, to support and further each other’s goals and mission. Specific areas where the two groups can work together include: standards-development tools & methods, mechanisms for obtaining industry feedback, international collaboration and outreach, an improved environment for corporate standards, growth of entity-based standards, and promoting critical mass in design automation related technology sectors.

The DASC updated its policies and procedures to allow for both individual and entity memberships. This became especially interesting (challenging?) when it came to voting rights. Could a person vote as both herself/himself and also as a company representative? Wouldn’t that be cheating? The DASC discussed the issue and decided to amend its policies and procedures to clarify that a person is not allowed to vote twice. Further, since voting privileges are earned in part by attending meetings, a person representing an entity at a meeting will still get credit for attendance as an individual.

Several entity-based standards projects for design automation are active under DASC sponsorship. The 1801 standard for low power IC design and verification was completed and approved by the IEEE Standards Association Standards Board in March 2009. The third update of the 1800 SystemVerilog standard – the first entity standard in design automation – has completed its initial ballot. The P1685 IP-XACT project which was initiated by the SPIRIT Consortium is expected to enter balloting later this year. The study group investigating the Verilog standard for analog/mixed-signal plans to start an entity working group by year’s end.

There is a small debate brewing over whether current mechanisms are well-coordinated or not for industry feedback on what standards are needed for design automation. The question was posed to the DASC to see if it wanted to set up a sub-group for coordination of and creating a roadmap for standards. The overall sentiment in the DASC was that it would be redundant with other industry efforts such as Accellera and SPIRIT – organizations that initiate standards based on industry need. I feel strongly that a standards roadmap effort would be fruitless and a waste of valuable time and energy. We tried this before and it was a failure. Repeating the effort would support the urban definition of insanity – doing the same thing and expecting different results. I also believe that the current mechanisms for developing new standards are working well. When a need for a standard arises, interested parties get together and form a working group under an existing organization (preferable) or under a new one. Donations are sought and work towards an industry standard begins. There is solid cooperation among organizations, especially between the IEEE and others such as Accellera, SPIRIT, OSCI, and VSIA (when it was active).

DASC is taking the right approach, IMHO, in that it will have a standing agenda item to invite representatives from other industry groups for input and discussion. This will encourage coordination better through this less-formal channel than a roadmap or additional, unnecessary infrastructure.

In my continuing series, “The 10 Commandments for Effective Standards”, the 6th commandment is:

Leverage Existing Organizations and Proven Processes

Standards begin in a variety of ways (a topic for a different post). Regardless of what motivates a standards effort, it’s important that a process is followed to ensure a successful outcome. When I say “process”, I actually mean all the ingredients required to produce a useful standard. These include: bylaws, policies, procedures, guidelines, rules of conduct, committee and governance structure, administration, websites, information management tools, PR, market acceptance, and formal accreditation. (Whew!)

Developing a new standardization environment is a monumental task. I’ve seen standards groups take 6 months or more just to get the basics in place. The best approach you can take to launching a new standards initiative is by doing so under an existing standards body – if possible – instead of creating an entirely new organization. Taking advantage of what an existing group has learned and refined over the years is a smart way to start. Existing standards bodies can offer you administrative, technical, marketing, and financial resources to expedite the standards process.

In my industry, and I’m sure that in other industries as well, there are several standards-setting and standards-development organizations to choose from. The criteria I suggest for selecting an organization to work with is finding one that you trust. Not just the one with the most money, or the biggest staff, or the fanciest website, but one that resonates with your goals. One that has a proven track record of delivering standards that are in use today by your customers. And one that has a positive reputation in your industry.

If, for whatever reasons, you can’t find an appropriate organization and decide to start your own, it’s still a good idea to leverage existing organizations. Some are quite willing to share their bylaws and policy documents that you can use as a template, if not adopting them completely. Accellera and the IEEE Standards Association, for instance, have shared their policies with other groups to help them get started or improve their existing policies. Using existing principles lowers the risk of something going wrong during the process and certainly shortens the time for your standard to be completed.

Observing the 6th Commandment for Effective Standards can save you significant time and money – both of which we could all use a little more of.

Standards are all around us. They serve a role in just about everything we use, eat, sell, and enjoy. While I play the standards game primarily in electronic design, the fundamentals of standards pertain to all industries.

At the Globalpress Summit in San Francisco on April 1, 2009,  the Chair of Accellera, Shrenik Mehta of Sun Microsystems, spoke about why open standards are good for users and good for business.

For electronic design, standards allow suppliers to build and sell tools that interoperate with each other. Design engineers use open standards to describe their designs in common languages and formats which promotes design reuse. Both suppliers and consumers of electronic design automation products benefit from increased productivity and reduced costs as a result of using standards.

Here’s a short video where Shrenik summarizes how standards help everyone. The video was produced by a Taiwanese reporter from Compotechasia, which I think shows that standards have an appeal around the globe. Hint: If you don’t understand Taiwanese, skip the first 40 seconds.