What Would Joe Do?
Peggy Aycinena is a freelance journalist and Editor of EDA Confidential at www.aycinena.com. She can be reached at peggy at aycinena dot com.
IP Update: Buyers & Vendors Beware!
March 10th, 2012 by Peggy Aycinena
Even this deep into the era of IP and design reuse, it’s been my thesis that things are not quite as far along as many in the industry would like you to believe. With that attitude in hand, I spoke with three different companies in the IP space who disagree, although they admit issues still remain.
You can read my conclusions below from what they had to say, or you can read the original interviews and draw your own conclusions …
* Hal Barbour, CEO at CAST
Where things stand …
* IP is a reality: Over the last 10 years, the reluctance to buy IP has subsided, because third-party IP is better than ever, and the companies that sell it have come to see themselves principally as product companies, not services companies.
* Lots of different types of IP: Vendors are selling processor cores, standards-based busses, mixed-signal blocks, back-end design blocks, software blocks, drivers, foundation IP, etc., or any combination of the above.
* Standards and tools: Various wrapper standards and IP integration tools are easing the burden of using IP in a design.
* NIH still a reality: Concern still lingers, often without basis, that if I didn’t design it myself, I shouldn’t bank my product and my job on somebody else’s design bit.
* Risk Aversion still a reality: Buying IP is still not great for the highly risk adverse, people who need to guarantee a block is interoperable, meets required specs, and has been sufficiently deployed to work out the bugs.
* Internally developed IP still a reality: If an IP customer cannot find what he needs out on the market, he will develop it internally. Sometimes he’ll turn around and sell that IP commercially, but only in a minority of cases.
* A Triumvirate of Bad: Internally developed IP continues to suffer from poor documentation, poor support, and poor visibility across the larger organization.
* You have no choice: Time to Market pressures are driving people wild – and into the arms of the IP vendors who can help their customers avoid reinventing the wheel. Your competition isn’t going to wait around – they’re using IP, so you better accept reality if you want to meet your scheduling demands.
* Dearth of evaluation tools: There are still no standard tools for evaluating IP, comparing IP from one vendor versus another, and no automated procedure for partitioning a design to maximize what’s implemented in software, what’s implemented in hardware, and how much of each should be purchased off-the-shelf from third-party IP vendors.
* Judgment call: People use more nuanced judgments to decide which IP to use. Is it adequately documented, cost appropriate within the bounds of the project, of a high-enough quality, and well-supported by the vendor?
* Forums: There are numerous forums, and other resources, which offer info on the features and capabilities of various types of IP, but it’s still up to the customer to determine the worth and applicability.
* Internal groups: Most companies who buy IP have internal groups tasked with evaluating the stuff.
* Start with a plan: All decisions with regards to buy or build fall out of a) the product plan, followed by b) the market plan, followed by c) evaluation from Engineering. Without this pre-planning, nobody should be risking the tens of millions needed today for product development.
* Buy from the Big Guys: If you’re using IP that’s been widely adopted – stuff from the 800-pound Gorilla IP vendors – bugs and irregularities are minimized, thanks to feedback from a large customer base.
* Differentiation a problem: If you are using the same IP that everybody else is using – stuff from the 800-pound Gorilla IP vendors – differentiation can be a problem.
* Differentiation not a problem: The amount of IP in a design today is still small, so differentiation between end-products is not really a problem.
* Lazy Gorillas: Sometimes the 800-pound Gorilla IP vendors get lazy, and their products grow less competitive.
* Small IP vendors on their toes: This is a good opportunity for the small IP vendors to prove themselves, but they usually need someone within the customer organization to advocate for them, and for their products.
* Hybrid business models: IP vendors are using a combination of business models – one-time use fees combined with royalty models.
* No royalties: Small IP vendors are still relegated to competing on price, therefore rarely go with the royalty model if they want to compete with the 800-pound Gorillas.
* Royalties are easy: It’s easy to track IP usage during manufacturing, ergo the royalty model is a good one.
* Royalties are not easy: It’s not always easy to track IP usage during manufacturing, so only the big IP vendors go with the royalty model.
* Foundries in IP: The foundries are making their presence known in the IP area, selling libraries and assisting companies in tracking their IP – for a price.
* The O.K. Corral: There’s still somewhat of a Wild West aspect to the IP industry – some vendors are there for the long haul, some are not.
* Buyers Beware! Know that nobody’s ever 100% happy with their IP – complaining is a permanent part of the Human Condition.
* Vendors Beware! Know that nobody’s ever 100% happy with their IP – complaining is a permanent part of the Human Condition.
* Opportunities abound: There’s still a huge amount of opportunity for IP vendors, product developers, and everybody else in the semiconductor supply chain because the worldwide demand for better, faster, smarter, sleeker and cheaper continues relentlessly.
Tags: 800-pound Gorillas, Bernd Stamme, CAST, Hal Barbour, Internally developed IP, IP evaluation, IP integration, IP vendors, IPextreme, Kilopass Technology, Market Plan, NIH, O.K. Corral, One-time use, Product Plan, Risk aversion, Royalties, Time to Market, Warren Savage