Core Values

Here on a beautiful sunny spring day in California, I had the pleasure of visiting a few small wineries and tasting the wares of the local vintners. Peering through each glass of delightful ruby liquid, I found myself taking careful note of the characteristics of each wine. Did the wine have "legs"? How was its "nose"? Did the taste explode in you mouth like a cornucopia of fruit and flowers? The evaluation process was both complex and personal.

Perhaps buying IP has some common elements to that of selecting a wine.

Let's start with so-called Standards IP, like USB or PCI. These are the Cabernet Savignon and Chardonnay's of the IP world. Every winery bottles them as the public appetite for such standards is steady. However, every winery is different and their product varies dramatically even though the grapes may be genetically identically. In the end it's not the grape that is the differentiator it is the skill of the wine master whom can get it just right. In Standards IP it is the skill of the architect and the design teams which can take an industry standard technology and produce a differentiated product that "tastes good" to the customer. It goes beyond functional correctness of the product and more into more subtle areas like how easy is the IP to integrate, how good was the support, how easy was the company to do business with? A bad experience in any of those areas can leave a bad taste on customer's palate. (cracker please...)

Another sort of buying behavior is related to the prestige associated with the brand of the product you're buying. There is a certain kind of message you are sending your guests when you serve Chateau Lafite Rothschild Pauillac that is different from serving a Charles Shaw (also know here in California as "two buck Chuck".) When I worked at Tandem Computers, sales used to complain about customers who chose inferior machines because "nobody ever got fired for buying IBM." I suppose in the IP world the modern equivalent would be that no one is ever fired for buying ARM. Perhaps people have a soft spot for three-letter acronym companies. More likely, these companies have done a good job of making people feel comfortable that there is value associated with the big price tag. As a business professor of mine once said "People just want to pay more", explaining that given equivalent products people will pay more money for the one that makes them feel better about themselves after the purchase is over.

I feel that the IP market selects its winners very quickly. It indexes the players according to the quality of the products, the stability of the company, and the overall customer experience. Winners will command higher prices for this value and soon find themselves on the top shelf.

I would highly recommend to every single person living in the western world to read Thomas Friedman's The World is Flat - A Brief History of the 21st Century. No other text I've read in the last ten years has scared the hell out of me like this little tome. It features a little African proverb that haunts me every day:

Every morning in Africa, a gazelle wakes up. It knows it must outrun the fastest lion or it will be killed.
Every morning a lion wakes up. It knows it must outrun the slowest gazelle or it will starve.
It doesn't matter whether you are a lion or a gazelle. When the sun comes up, you better start running.

Friedman's book is essentially about the acceleration of competition and the emergence of India and China to take center stage in the 21st century. We see every major high tech company with a presence in these places. Universities are competing with centuries old institutions in the West for the minds which will dominate the decades to come. My son has a friend (with Indian parents) whom plans to return to India to attend the IIT rather than MIT.

But the West has a secret weapon it sometimes forgets about. We've got IP. Lots of it. And that IP gives us a very valuable thing--a head start. Let's take a look at what's going on in semiconductor today:

• 60% of a modern SoC is comprised of IP from other companies
• Fabs are increasingly non-differentiated and compete solely on cost
• IDMs are differentiated by their system level knowledge and use of IP

The fact that IP and system level expertise are the critical success factors in semiconductor means that the West has maybe a 10 year head start over its faster, more agile, hungrier rivals in the East. The question is whether that gap will be narrowed or widened as the next decade plays out.

I hated the Oscar-award winning film The Departed. Friends and family ask questions ranging somewhere between "why" and "what's wrong with you". Of course, nothing's wrong with me except I'm an engineer and can perhaps get distracted by the details. In the movie, Jack Nicholson plays a mobster who's infiltrated the police and FBI and can pretty much get away with trafficking contraband with immunity. One item he traffics is "microprocessor chips" (use your fingers like Dr. Evil for the right effect) sold to the Chinese in plastic 208 pin quad flat packs in a briefcase. Apparently Fry's was out of stock.

It seems the screenwriters haven't been paying much attention to the semiconductor industry's interest in selling as many chips to the Chinese market as they can. Instead they fell back into familiar bogey-man stereotypes from the Cold War era. Nonetheless, most IP companies are not yet so enthusiastic about letting their IP into China, especially in source code format.

As semiconductor industry sorts out the "new world order", one thing is clear--design is moving east and today, design means using IP. A lot has been written about selling IP to China and the dangers associated with cultures that might not respect the principles of ownership. The consensus in the IP community is that China is quickly moving in the right direction towards respect for protection of licensed IP, but still has a way to go. The feeling is that sooner or later China will have enough IP of its own that it will be concerned about Western companies stealing from them and there will be an "Aha! Moment." In the meantime, we are careful.

This transition period may serve a useful purpose for the industry. During DAC 2006, Synplicity put forth an innovative proposal to create an industry standard IP encryption technology that all of EDA could implement to permit a fully secure flow for designs that use IP. This would allow IP companies to deliver their IP as binaries with a key to customers that would allow EDA tools (and not the customer) to see the code and simulate and implement the IP as part of the design. Such a technology has long been the holy grail of the IP community which would simultaneously allow companies to protect IP and also develop new kinds of business models like subscription licensing which would provide more flexibility and lower costs for customers. The proposal has been provided to VSIA for standardization which is chartered with driving its adoption across the EDA community. I'm hopeful that something useful pops out of this effort.

In the meantime, we must stick with legal contracts, secure design centers, watermarking, and other "trust but verify" methods. Even if the EDA industry comes through with the encryption technology we all need, I'm sure Hollywood will also be slow to pick up on the news and we'll surely see a movie sequel with mobsters passing a briefcase of "RTL Source Code" printed on green and white fan-fold paper.