Portable wireless product makers are asked to design and build 30 to 50 new products per year-"products" being anything from simple software changes to major new device platforms. And they have to service various combinations of all of the different air-interface requirements in the world, from GSM and AMPS to CDMA and TDMA-with 3G and W-CDMA coming up quickly.
Those vendors face intense pressure to come to market with new products that meet the needs of the consuming public at a price point the service providers can support. If a product maker is as little as one to two months late to market, the contract may be lost altogether and a new design cycle must be started to meet the next window of opportunity.
The EDA industry has not been able to keep up with the needs of all of the commercial wireless product designers. That is because the economics of advanced EDA tool development does not permit each EDA company to develop each of its customers' wish lists without advanced development dollars. And most OEMs are not willing to make such investments in tools in advance of delivery. This is a serious catch-22 situation for the design-automation industry.
Time-to-market can be significantly affected at the design level and, more specifically, by the design station. The designer has to have the design tools, knowledge and accurate data readily available to make materials and manufacturing trade-off decisions, which will affect manufacturing, yields and first-pass design successes-not to mention time-to-market.
One of the tasks of the International Wireless Packaging Consortium (IWPC) is to try to unify and simplify the EDA-tool wish lists of the wireless product makers, focusing on packaging and interconnect technologies. The IWPC's Electronic Product Design System initiative (EPDeS), launched in October, brought together wireless-industry leaders from all levels of the supply chain.
To reduce time-to-market as well as costs, the manufacturers focused on several issues. For example, there is no common language to communicate advanced packaging technology (materials data, manufacturing process design rules and models) among all layers of the supply chain.
Another issue is reducing the "vendor-dance" time. It takes 10 to 20 times more calendar time to design with a new material and/or a new manufacturing process than to use the same tried-and-true materials and manufacturing processes-even if that new material or manufacturing process promises lower manufacturing costs or faster time-to-market. That is due, in large part, to the "vendor dance"-the amount of time it takes a designer to find out about and collect sufficient and accurate information to do a new design.
A major time-eater and error-prone challenge for RF designers is the inability to transfer files among different or special-purpose mechanical, electrical and thermal EDA tools, without having to redraw the design. The redrawing process adds a significant amount of time to the design process and is fraught with the possibility of introducing errors in design entry. Those may not be caught until the late-prototype or manufacturing stages, at which point they are very expensive and time-consuming to fix.
Another conclusion of the IWPC was that 3-D RF design tools are too slow, especially 3-D electromigration-analysis tools. It often takes less time to actually build and test a prototype than it does to iterate enough 3-D simulations to converge on a suitable solution.
The goal of the EPDeS initiative is to provide the future designer with everything he or she needs, on the desktop, to reduce time-to-market and cut costs of RF designs. This vision, articulated by Paul Draxler of Qualcomm, puts the RF designer at the center of the RF design environment, rather than at the mercy of supposedly "integrated" EDA tool-systems topologies.
A further goal is to enhance commerce at all levels of the supply chain by creating a much tighter link between suppliers and users, using efficient Web-based links that dramatically slash the time needed for the vendor dance-for example, getting accurate and timely vendor data to answer a specific design question and have that data readily usable by the design tool.
While those are lofty-some say unattainable-goals, through interactive discussions IWPC members have distilled the wishes of the OEMs to a few tasks:
- Create a vendor-controlled, Web-enabled materials database system to provide complete and accurate mechanical, thermal and electrical data.
- Develop a vendor-controlled Web-enabled design-rule database system, accessible by properly enabled EDA tools.
- Create a Web-enabled models database.
- Develop a means for EDA files to be transferred among mechanical, electrical, thermal and database tools. IWPC members are working to define what is called a "neutral dynamic hub" concept to meet that end.
- Support EDA tool suppliers so that they can read, via the Web, the EPDeS materials database, design-rules database and models database, whose files can be transferred from one EDA tool to another using the neutral dynamic hub.
- Build an EPDeS Web site that will act as a portal to bring together EPDeS Web-enabled data and EDA tools.
Once those tasks are operational, commerce will be enhanced at many levels. OEMs can build products with faster time-to-market and at lower manufacturing costs, due to more accurate materials, manufacturing and models data that provide the opportunity for higher first-pass manufacturing yields. Materials and circuit-maker suppliers can dramatically reduce the cost of finding new sales prospects and communicating complex information to their customers and prospects via the Web. EDA tool suppliers will be able to grow their "marketplace pie" through enhanced use of design tools to facilitate communication at deeper levels in the supply chain.
The IWPC has organized several working groups to think through and advise the development process to meet the goals of the EPDeS initiative. The working groups cover materials database and design rules, CAD file transportability, Web-site development, models and funding.
The IWPC is planning its next EPDeS workshop in the November time frame, when an IWPC-funded demonstration of a prototype electronic product design system will be made.