I suspect that in another year we’ll all stop talking about OpenAccess (OA) like it is something special and treat it the way it should be, that it is just another database. Having said that, I know I’m going to get plenty of email about my portrayal of OA from colleagues and others but that is the way I see it.
Let’s not fool ourselves, today OA is a big deal because it truly is a different way that will allow designs and design information to be portable between applications from the same or different vendors without the need import and export data that loses intelligence in the translation. It is as simple as that to the users.
The motivation behind all of this is to have an “open” standard for IC design information and to get it adopted within the EDA industry. These goals are defined by the OpenAccess Coalition* which seeks to provide its partners with technology and support that will:
Provide tight integration and incremental design flows using design tools and data from multiple sources
- Ease integration of internally developed tools with those from EDA suppliers
- Provide more cost effective technology transfer of university research into the design flow
- Simplify technology sharing for collaborative development between business partners for design tools and design data
It is intended for industry-wide interoperability where OA will permit companies to create flows utilizing design tools from different EDA vendors to best suit their individual and custom needs. Cadence, along with others, is moving forward with OA. Some have announced their intent to support and use it, while there are plenty more poised to do the same. I foresee that in the next twelve to eighteen months we’ll see more EDA companies accessing design data and provide interoperability via OA. I think a bit of history will help put some of this into perspective.
As reported in 2002 by my colleague here at Cadence, Richard Goering, OA Version 1 database code was released to the coalition in February 2002 by Cadence. It was originally known as the Cadence Genesis code which was intended for digital designs. Later in June of 2002 Version 2 was released and it now had support for analog and mixed-signal designs along with technical features to facilitate support by developers. OA was now fully in the public domain courtesy of Cadence. At that time there was only two other members of the OpenAccess Coalition. The OpenAccess Coalition has come a long way since then.
One last thought before I wrap-up. There have been some recently noted comments saying that OA is only good enough for custom IC design but not digital. How can this be?
Didn’t I just get finished saying that it is “truly is a different way that will allow designs and design information to be portable between applications”. And didn’t I quote the OA Coalition in saying that this would, "Provide tight integration and incremental design flows using design tools and data from multiple sources". We, Cadence that is, took “multiple sources” to mean from different tools addressing differing design domains.
I beg to disagree with those making this claim. OA is a design database for representing information regardless of the design type. Our digital tools prove it. Encounter reads OA and writes OA. Virtuoso reads OA and writes OA. Virtuoso exchanges information with Encounter via OA. OA carries both the analog and digital design information.
Back to my premise which is OA is big news now, and Cadence along with others will talk it up but it will fade into the background and simply be the database of design going forward. I like to think that it will allow you to “Create, collaborate, and just build it.”.
*Bulleted items are taken directly from the OpenAccess Coalition website found at www.si2.org.