Thursday Sep. 11, 2008
One on One: Brian Gardner
Brian,
Its great that you had time to talk with me about some questions I'm asking leaders in the IP community. Let's get started.
1. As an IP company, what is your #1 challenge you have with your customers and what are you doing about it?
Fit: meaning having what the customer wants exactly. The fundamental conundrum in IP is the that the IP supplier wants to build one thing and sell it many times, while the IP buyer wants to buy something that fits perfectly into his system without modifying it. Modification requires time, resource and risk, and greatly reduces the net value of the IP.
At Denali, our answer is configurability, and our controllers are compiled. Customers select via GUI from dozens of options and create a metadata file that drives the compiler. Custom RTL, user documents, scripts, and test files are created and all these are thoroughly regressed. Most times, customers never modify the code. Also, the average customer tunes his controller as the design matures, ordering 2-3 different configurations over the life of the project.
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?
Quality and fit, as discussed above. On quality, customers should be much more demanding—they don’t expect enough—and be willing to pay for it. On fit, they should expect better fit… BUT, they need to demand that their engineers strive to use standard blocks. Just like engineers now expect to use standard GDSII cells, they should expect to use standard IP blocks. Stop trying to sub-optimize.
3. The IP business model undergoes periodic attack from the media, do you think the IP business model is broken?
I’d say that it’s just not mature. Even though IP has been around a long time, I’d say the “tipping point” has yet to occur. I’d speculate that, if we look at total chip unit volume, there are no categories of IP that have crossed the “50%-is-third-party-IP” line. Even in standard cells and compiled memories and CPUs, I’d bet there are more units with proprietary IP than third party. So, we just aren’t there yet.
The real change will happen when chip business executives, the GMs who are accountable for the bottom line, are starting to worry about the problem. When it’s no longer a VP or Director of engineering make-or-buy decision, but rather a critical business initiative to use IP. When using IP better than your competitor is seen as core competence.
4. The IP market is dominated (revenue-wise) by a few large players, yet there remains hundreds of small outfits. Why is that?
It’s a sign of the immaturity of the industry. Once it becomes a business issue, and the requirements to play are clear, the small outfits will either fade away or focus and win. I do think, though, that IP should be an area of innovation where startups can flourish as long as they have vision and focus. The number of chip startups is fading as the costs get high, so I see the IP industry as the next frontier for the entrepreneurial high tech folks.
5. Should there be more or less IP companies?
I think I answered it above. I’d love to see small companies flourish!
6. What role should EDA companies play in the IP market?
As an enabler, enabling IP to be developed cost-effectively and used easily. Creating standards. Whether EDA companies should have IP divisions is up for debate. On one hand, as a buyer, it’s nice to have one-stop shopping and guaranteed tool integration. On the other hand, IP and EDA are not similar and require different channel attributes, different development disciplines and culture, and different support. They don’t fit too well together, and can create a tricky management problem for the company.
7. What role should Service companies play in the IP market?
They can fill the gap between IP suppliers and the SoC manufacturers. They can put together the IP, and modify as needed. They can implement or harden elements. They can act as a channel… especially, the backend services or fabless ASIC houses.
8. To what extent do you think the IP industry has come to grips with its quality issues?
Individually, each supplier has done what it takes over time because the cost of failure is so high. As a group, there is work to be done.
9. Is there a need for greater standardization in IP?
Yes, but it’s not a panacea. IP fit will require compromise on both sides: better use of standard IP and better IP configurability.
10. Ten years from now, what does the IP market look like?
It’s likely to be the forefront of technical innovation… where every EE wants to be. As SoC design becomes increasingly a job of integration (I’m leaving firmware aside for now), the real innovation will come in the new IP blocks and subsystems.
Now, on the other hand, I’m waiting for the promise of nanotechnology and $10M desktop fabs where “chips” are grown like DNA. And, moving above RTL, to the powerpoint compiler where huge SoCs are compiled from 20 slides. And SW is something the customer writes. Then chip companies will rule again!
Posted by Warren Savage on Thursday Sep. 11, 2008
|
Add a Comment
|
|
RSS Feed
Add this blog to your RSS newsreader!
Archives
Meet the Bloggers
One on One: Jack Browne
One on One: Kathy Werner
One on One: Ralph Von Vignau
One on One: Stephen Padnos
One on One: Johan Van Ginderdeuren
One on One: John Koeter
One on One: Hal Barbour
One on One: Joe Abler
About the Author
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.
|
