Semiconductor IP News and Trends Blog
Part II: Stop That Thief! – IP in Global Markets
Expert panelists from Globalfoundries, IDT and the legal community share their experiences on many topics, from the use of leading edge and middle-edge IP technology to the dangers of outsourcing IP development and, finally, to the legal struggles in emerging countries. This is Part II of the ongoing coverage from the DAC Pavillion Panel on IP security and quality issues in global markets.
Last time, I reported on Adam Traidman’s opening remarks as moderator for the DAC Pavillion Panel on IP quality and security issues in global markets. Now, I present the follow-on remarks from the panelists, including: Gerd Teepe from GLOBALFOUNDRIES in Dresden, Germany; Alan Coady from Integrated Device Technology (IDT) in Ottawa, ON, Canada; and Jonathan T. Kaplan from the Law Office of Jonathan T. Kaplan in Vancouver, WA. — John
Traidman: What about the technical, business, and legal aspects of cross-regional IP sourcing? What can the IP ecosystem do to promote business integrity worldwide?
Teepe: IP integrity is based upon trust. You must trust that your IP works, that it comes in time and that it doesn’t jeopardize your business. If the IP is from a questionable source, it can entirely upset your business. On the leading edge of technology, there is one major problem. It is not an integrity issue, but rather one of priority in terms of time-to-market (TTM). Quality is linked to TTM, since bad quality will blow your schedule and your TTM will increase. So you need trustworthy, reliable IP partners.
This is why there is less of an IP problem in leading edge technology. Where the IP integrity problem exists is in mainstream IP, in the middle-edge. It is hard for companies just entering the IP market, such as those providing MP3 technology. This middle range is a grey zone which we need to watch.
Coady: I see three camps for the IP reuse engagement model. The first camp is off-the-shelf IP, meaning that the IP is mature and has been well-used. This model is the least risky, most deterministic in schedule and quality. The second camp is the midstream IP, which is IP that contains new qualities and features. Designers want to use this IP to be in the forefront of new designs. In this area, you have other risks and considerations.
The third engagement model is one of outsourcing IP development to handle expertise that you might not have in-house because of resource constraints. You might want to own this IP in the end, or you may want to just seed the development of it and pay a portion of it – but not the entire development amount. You’d be the lead customer of the IP, defining it and the requirements.
Each of these models has unique benefits and challenges which I put into two additional camps. One is an engineering or execution camp. The other is a business or legal camp. You need be “on your game” in each of these camps since either on can burn you. Obviously, off-the-shelf technology is easier from an engineering or execution point of view. Off-the-shelf IP may also be easier from a legal view, although you still need to worry about infringement, escrow, indemnification and others.
Maturity is another issue for this camp of benefits and challenges. Off-the-shelf IP doesn’t necessarily mean that the IP is good. The IP might be good in the sense that it has been used in many devices, but your device may utilize the IP differently – in a different system or context. You run into other issues. Conceptually, that is why off-the-shelf is the easiest approach but it is not risk free.
Then you move down to mid-stream technology. The developers are creating the IP on specification, but you are engaging with them early on. You might not have too much say in the requirements but you want to be one of the first customers. Now your main concern is one of execution, in addition to all the other issue from the off-the-shelf camp. For mid-stream technology, you need to have a better understanding of the company and the quality of the engineers.
The next camp is outsourcing the development of the IP. This is an exaggeration of the previous camp. Here, you are really interested in the company and in the quality of the engineers. You meet often with the company and the designers. You actively drive the requirements in certain ways. But you run into unique problems when you deal with emerging companies, which deals with the way in which we communicate. In North America (NA), we have unspoken understandings when we communicate with one another. But I’ve dealt with IP companies in India and China and the assumption that we make in North America are not the same ones that they make. In NA, we fill in gaps for what we don’t say. They don’t fill in those gaps, so you have to be very specific in dealing with India or China. Also, you have to provide very specific directions. If you tell them to do the wrong thing, they will happily do it. This is yet another level of concern.
Also, people in China and India do not stay at companies as long as in NA. Depending on the engagement model, you are banking on them getting in done in time, but the team that you start the project with may not be the same team that finishes the project. That doesn’t help time to market or quality.
Further, if you are specifying some IP to own, you sign a Non-Disclosure agreement (NDA). But what does an NDA mean? The NDA is with the company. While the NDA can get back to the people, this is seldom the practical case. You don’t sign contracts or do anything with them except to say “please behave.” If you have to enforce a contract, then you have to spend more money. So these are practical issues.
Another one is the maturity of the team. Certainly, there are good people in India, but there are a lot of junior people, too. I won’t use up any more time, but these are the things that I see as considerations.
Kaplan: I’ll jump in now with some legal perspective. I’m going to focus on China and India. Many business authorities advocate taking a joint strategy of China and India, not China or India. Also, the better view is that you don’t go to Asia to save money. Instead, you go as part of a long term strategy.
These are emerging markets which are undergoing rapid economic and industrial development. From a legal point of view, there are basically three stages of a law. First is the promulgation of the law, which is when it is originally passed. All you have is something on the books.
In the next stage, you have the institutionalization of the law, when the country creates institutions to enforce the law.
The third and final stage – assuming everything goes well in the institutionalization stage – is the internalization of the law. Here, the law is essentially self-enforcing with voluntary compliance. Everyone in the country knows the law and thinks it is ok or knows how to live with it. People follow the law without needing enforcement.
The emerging market countries, certainly China and India, are in the institutionalization stage of the law. This is the most critical stage of legal development. I think the question with emerging market countries is whether they will make it through that critical institutionalization stage. Key concerns are that the enforcement must be consistent, transparent and is seen as procedural fair. These are all things that both countries are working on.
There are interesting comparisons and contrasts between the two countries. One is an authoritarian country, while the other is basically a democracy. But they both have this strange dichotomy where there is a certain level of what might be called “lawlessness.” This happens because of the enforcement issue. There are many other issues that I might cover, but this seems like a good place to stop.
Adam: You hear that big IP companies don’t want to do business in these emerging markets because they are afraid of the GDS being copied or the RTL being sold on CDs on the streets of Shenzhen, China – as what seems to happen with many of the EDA tools. Also, the licensing of the IP, from a technical perspective, is more difficult. At least with the software and EDA tools, there are mechanisms in place to safeguard the license. But for the silicon, you almost have to pull back the package and try to scan the chip to see if your IP is there. It is very challenging.
Next time, look for a variety of interesting questions from the audience attending this panel. — JB