On Cores
Meditations on the semiconductor and IP industries
By Warren Savage, CEO, IPextreme
Tuesday Sep. 09, 2008
One on One: Bill Martin
Bill,
Thanks for taking the time to talk with me. As a long time IP guy, I'm very interested to hear your opinion about a few questions.
1. As an IP company, what is your #1 challenge you have with your customers and what are you doing about it?
The largest issue we faced with our customers was integration issues. These issues arose from multiple reasons. Some of the reasons were: trying to modify purchased IP for additional features/capabilities; a lack of specification knowledge; ambiguous specifications; time pressures; lack of larger solutions, etc. To help remove/ease our customers’ integration issues, we started to release compliance-tested subsystems that provided a larger solution that included digital, analog and embedded drivers.
2. Putting the shoe on the other foot, what do you perceive as your customer’s # 1 challenge and what should they be doing about it?
NIH. In my early engineering days, I always wanted to ‘tweak’ designs I worked on. Customers are striving to optimize for area/performance/power and provide additional value, but with the integration capabilities offered today, along with the tight time to market (TTM) windows AND reduced resources, customers will need to increase IP reuse (SIP, Embedded IP, VIP) inside their designs.
When any changes are performed on IP, whether internally designed or purchased, the value provided by the creator has been wasted and much of this work has to be redone. This re-work would entail: design, verification, validation and compliance testing. If IP modifications are required, designers should try to keep these external to the purchased IP in order to maintain the value purchased.
The industry needs to adopt tools and methods that help users to understand the quality of their IP. If purchased IP is not compliant with industry specifications, hold the IP suppliers accountable for correcting their products rather than forcing users to correct the problem in their own design. The best correction is at the root cause.
Yes, this sounds idealistic. You might be helping a competitor with their product which contains the same purchased, buggy IP. Customers need to realize that they could be ‘bit’ by the same behavior. The Olympic Games were just completed and one US swimmer demonstrated how we should all act. Dara Torres asked one of the pool officials to temporarily postpone her 50 Meter Freestyle Semifinal Heat while one competitor needed time to change a defective swimsuit. Dara could have easily done nothing, eliminating one of her competitors and improving her odds to win. Dara did not. She felt she wanted to race against all competitors and had confidence in her product (herself). No one could ask questions after the heat since all competitors swam.
3. The IP business model undergoes periodic attack from the media; do you think the IP business model is broken? Please elaborate.
Everyone states they want a consistent business model, but in dealing with many customers, they all had their own business and legal requests based upon their specific situations. The market will dictate what models will be used as well as which suppliers survive.
4. The IP market is dominated (revenue-wise) by a few large players, yet there remains hundreds of small outfits. Why is that?
There are probably two key reasons: First: many of the large players started years ago and created ‘STAR IP’ that could command premiums, royalties and few (if any) competitors. ARM is a very good example. ARM created a small, power efficient RISC core that was targeted towards high volume consumer applications. Second: scale. You either need to be a very large or very small IP producer. Small companies tend to focus on a few niche product lines and have lower overhead rates that are not subject to corporate infrastructures. Small companies are able to survive on lower revenue streams. Large companies have a wider product portfolio and produce revenue streams that offset corporate taxation. Either large or small companies can innovate rapidly. So for the ‘mid sized’ firms, it is a question how quickly you can grow the business to remain viable.
5. Should there be more or less IP companies?
More.
Healthy markets have more suppliers that offer product breadth (more standards offered) as well as product depth (more product options within a supported standard). Healthy markets are not built on monopolies that can hinder efficiencies or innovations. More competitors force you to continually innovate in your business models as well as your products.
6. What role should EDA companies play in the IP market?
EDA will always participate in IP markets. Their tools are used to design and verify the original IP and used by IP consumers to integrate and verify the purchased IP into their designs. If we consider IP to include digital, analog, verification, and embedded software IP, many of these are tied directly to various tools and methodologies provided by EDA companies. In addition, it would be interesting to look at ‘classical’ IP providers and see how many of them provide some type of tools. Today, the demarcation between various solution providers (IP vs. EDA) is a blur.
7. What role should Service companies play in the IP market?
Service providers are part of the development solution in addition to SW tool and IP providers. They can be the experts that perform specific functions during the development process.
8. To what extent do you think the IP industry has come to grips with its quality issues?
Good question.
I know that we had some quality issues that we addressed and resolved. We continually monitored bug trends for each product to ensure that our quality continually improved. If an error was found, we determined the root cause so that similar errors were not overlooked in our development/release process. This does not mean that we did not have service requests (SRs) logged by our customers. As I mentioned earlier, many of these SRs were integration issues or specification interpretation questions with a small percent that turned into actual bugs in the IP.
I do not think this is solely an ‘IP supplier’ issue. It was very rare that we ever received customer RFQs that asked for HIP or QIP ratings. Customers were not demanding this from us nor considered this in their purchasing decision. Wider adoption of these tools would send a clear message and allow everyone a better understanding of high quality IP’s value. IP purchasers need to understand that short term gains (lower IP acquisition costs) might be more costly in the long term. These costs would include: additional resources to re-engineer, additional mask/wafer respin costs, missing market windows, etc.
Even though these tools, HIP and QIP, had some built in subjectivity and were completed by IP providers, they do offer standardized questions for any IP block considered for purchase. See additional comments in question 10.
9. Is there a need for greater standardization in IP?
Yes and no.
Yes, we need to have standardized bill of materials (BOMs) that allow customers to get all of the data required for them to successfully adopt and integrate IP into their design.
Maybe, if we are discussing standard (functional) specifications. Each industry group (IEEE, USB IF, etc) determines and approves the required and optional functionality for a given specification. IP suppliers will use this as a basis for their product line which may offer variations that all fit within this specification. Similar to other consumer products (i.e. MP3 players, cell phones, etc.) that offer multiple options and pricing to suit various consumers needs. If the industry wants greater IP functional standardization, then this will need to be driven by the various industry bodies that own these standards.
Yes, if we are discussing how IP is compliance or certification tested. Similar to design processes, these need continual improvements to keep up with technology and requirements.
10. Ten years from now, what does the IP market look like?
I do not expect revolutionary change.
It will be evolutionary as the community adopts new business and technical standards. We will continue to see the very large and small IP providers while the mid-tier suppliers will either grow or acquire to be a large player, be acquired or go out of business. The same can be said for service providers.
More IP suppliers will migrate to subsystem solutions to help accelerate their customers’ successes by removing various levels of integration (and interpretation) issues. Customers will demand this type of integration given the integration capabilities that will exist. This will dramatically lower the cost of support and integration.
An independent quality lab is established that uses HIP/QIP-like tools, in addition to other tools to test all IP, could have a yearly subscription or per-test fee for IP consumers and providers --similar to Underwriters Laboratories, focused on quality and rather than safety.
Posted by Warren Savage on Tuesday Sep. 09, 2008
|
Add a Comment
Warren Savage, President and CEO of IPextreme, is a well-known and published authority in the field of semiconductor intellectual property.
He has a long history of pushing the envelope of design methodology from his work in fault tolerant computing at Tandem Computers in the 1980's and driving reliable design metholologies into commercial practice at Synopsys for its DesignWare IP product in the 1990s.
Much of his thinking became embodied in the seminal book on IP reuse, the Reuse Methodology Manual. Warren is taking his vision to the next level with his latest company, IPextreme, which is focused on enabling broad commercialization of IP captive in large semiconductor companies.
1. As an IP company, what is your #1 challenge you have with your customers and what are you doing about it?
Warren Savage, President and CEO of IPextreme, is a well-known and published authority in the field of semiconductor intellectual property.
He has a long history of pushing the envelope of design methodology from his work in fault tolerant computing at Tandem Computers in the 1980's and driving reliable design metholologies into commercial practice at Synopsys for its DesignWare IP product in the 1990s.
Much of his thinking became embodied in the seminal book on IP reuse, the Reuse Methodology Manual. Warren is taking his vision to the next level with his latest company, IPextreme, which is focused on enabling broad commercialization of IP captive in large semiconductor companies.
