May 14, 2007
Patent Licensing – MOSAID Technologies
Please note that contributed articles, blog entries, and comments posted on EDACafe.com are the views and opinion of the author and do not necessarily represent the views and opinions of the management and staff of Internet Business Systems and its subsidiary web-sites.
Is there any significance to the company name, MOSAID?
From my understanding it is CMOS, NMOS, metallic oxide semiconductor. AID is help. So it is help with semiconductor chips.
MOSAID is publicly traded. What is the annual revenue? How many employees?
The number of employees underwent quite a bit of change in the last few months. In February we had around 120 people. After the sales of the systems business, we were down to about 60 people. If we are successful in selling the semiconductor IP business, we will be closer to 35. It has undergone quite a bit of transition.
I do not have the figures in front of me.
Editor: I can get them.
MOSAID 5 Year Financials
On a geographic basis Korea made up around 65% of revenue the last two fiscal years, Taiwan 15% and the US 7%. In prior years Japan had been the leading revenue contributor. The Systems division that was recently sold accounted for ~22% of total revenue.
Does the fact that MOSAID is a Canadian company have any benefit or downside?
I don’t think so. Not as our current strategy or focus is on patent licensing. Discrete R&D is another piece. We have an engineering R&D effort focused on flash interface technology that we think is very promising. We will continue that effort even after spinning off the semiconductor IP group. With these areas of focus the geographic location of the headquarters does not have a bearing on the ability of MOSAID to succeed in this market in a number of places.
The nice thing about Ottawa and not many people outside of the semiconductor patent world realize this, is that Ottawa is the reverse engineering capital of the world. MOASID actually gave birth to this industry. Along with design services they also provided for hire by-the-hour reverse engineering for other companies existing chips. They used that for competitive data gathering and also to figure out whether they infringed. When I was at TI, we utilized their services in the late 80’s. That was my first trip up here. Very shortly thereafter they decided to discontinue that business. They found it very difficult to sell their design services and testers to companies that would see
the MOSAID logo on infringement reports by a patent holder. The customers did not like that service. They spun it out. That gave birth to Semiconductor Insights, a very famous reverse engineering house here and that gave birth to ChipWorks. A third firm is GIF. The three of these are world famous reverse engineering houses that are used by companies all over the world to read back circuits and process steps on semiconductor devices for use in trials and licensing negotiations. Having that in our backyard is actually a big benefit. We have hired several employees away from these firms to help us in house with our assertion program.
MOSAID recently announced a deal to acquire some wireless patents.
That was form Agere Systems. It was a very competitive auction process. They initially sent letters out to something like 80 business entities from corporations to privately held organizations including licensing focused organizations, manufacturers, law firms, and investment houses. Just a variety of different people who might be interested and offered them to bid on this portfolio that had been collected from among their vast portfolio of patents. The source of the patents was a group out of the Netherlands that had worked for NCR. NCR was working on some technology to enable cash registers to be able to talk to one another and talk to a mainframe in the back of the store. That developed into the formation of 802.11 Working Group founded by these people. That was bought by AT&T who continued to invest in it. That was spun out into Lucent and then transferred to Agere. That was the lineage. These patents are fundamental. They crafted the portfolio selected it very carefully with lots of breadth. It covers 802.11 and also the base standards A, B, C, E, G and N, all of the different flavors. It also covers 802.16, the WiMax standard. The parameters for the bidding
process were that they wanted a substantial amount of cash but also a percentage of royalties that would get collected after the fact. That was the structure. There were several rounds and we came out on top, thank goodness.
Editor: Under the terms of the agreement, MOSAID will pay Agere US$25 million during MOSAID's fiscal 2007 fourth quarter, US$5 million in the fiscal 2008 fourth quarter, and eight subsequent annual payments of US$5 million, creditable against Agere's 20% of royalty revenues earned under a revenue sharing formula.
Editor: On May 8, after this interview, MOSAID announced a 10-year patent licensing agreement with LSI Corporation. Effective immediately, MOSAID has the exclusive right to sub-license more than 50 LSI patents relating to memories, DSPs, microprocessors, ASICs, and semiconductor processing technology.
When MOSAID licenses a patent, do you provide any technical assistance or transfer any technology or is it just a license to develop or manufacture something that is covered by the patent?
There are two different ways, probably more than that. For our purposes there will be two models. A naked patent license where there is no technology transfer, no knowledge teaching that goes along with the license. It will simply be an after the fact license. You are infringing on this patent. You may have independently designed it but we have the rights to enforce this patent against you because the inventor came up with the concept and protected it prior to your distributing the product that now uses it. So therefore, you now owe us a royalty. The other way we are
interested in pursuing based upon the research we are doing in our R&D group is to do both. To do a technology transfer, a know-how, a co-development with a manufacturer. A patent license goes along with that.
How do you know someone might be infringing on your patents? Do they come to you and attempt to negotiate or do you go out and try to find someone that is infringing?
It can happen both ways. Occasionally there is a proactive company seeking to resolve any difficulties with patents by approaching the holder of the patent first and negotiate a license. But more often than not it is done the other way around where we actually look at product announcements, particularly with respect to the wireless patents. If they announce a product that complies with a particular standard in which we believe we have patent coverage, then we will approach them and offer a license to the companies we believe need one. We also have for our DRAM portfolio very fundamental patents that don’t necessarily relate to any standards. But we think they are the best way of commercially producing and design DRAM. So we have an active reverse engineering program. We simply buy up products as they come onto the market from various companies at different densities and in different configurations and pay for the reverse engineering of those products to readback the circuit and compare it to our claims. Then we build claim charts and show them the evidence of infringement and negotiate or litigate if necessary. Currently we are in litigation with Infineon,
Micorn, Powerchip, and ProMOS in the eastern district of Texas.
If you suspect a firm of infringing one of your patents because the patents are fundamental and you send off notification to the offender, in general would that company be able to easily tell if they are in fact infringing?
This is a very case by case basis. Clearly, with the portfolio we have purchased from Agere that is the case. All we would have to say is here are some patent numbers and their engineers who are responsible for the product could read the claims and realize that there is a match with the standard and therefore they need a license. When it gets into very circuit specific patents and some process related patents, it is possible that they could easily do their own assessment with the identification of the patent numbers. But sometimes it takes a claim chart and an element by
element comparison to point them in the right direction to see whether there is infringement or not. Obviously there are different ways to interpret different claim elements. I think it is very helpful to provide the kind of guidance as to how we are interpreting the claims and how we are reading against their products.
Patent litigation seems to take a considerable period of time and money. Is it more a way of negotiating over price? Does each side typically truly believe they are right?
I think there has been a lot of press with this situation of patent trolls. I think the debate should not be over characterizing the patent owner as to whether they are a producer or not a producer but everyone should have the same interest in increasing the quality of patent system in the US. There is a very strong argument that companies should stand up, fight and take things to litigation and not pay off nuisance values, if the patents at issue are not solid patents. I certainly understand why companies go to litigation. They want to send a message that I am not going
to be held hostage by invalid patents that are not infringed solely to pay them off so I can avoid litigation costs. That’s a positive development for all patent holders and companies that we want to respect IP on both sides of the fence.
But there is a fair amount if not the majority of litigation that occurs because it is a price negotiation, a leverage point. Lots can happen in litigation in terms of the interpretation of the claim language and how it reads and it is a jury system in the US. So there is some gamble to be taken. Sometimes just filing litigation and getting through portions of the process can encourage one side or the other to reduce or increase its offer on the table and settle. The vast majority of cases of course settle prior to trial.
How does a patent owner come up with a proposal for fees and royalties? It is related by the original cost to develop? Is it the potential value to the licensee?
In the case of semiconductor patent licensing the industry is very mature. The acceptable range of rates that are possible to get on portfolios is pretty well established particularly at MOSAID. We have something like 75% of the world’s DRAMs are already licensed to our patent portfolio. We know what an arm’s length transaction should bring in terms of the value to the market for a license to a patent. But licenses can be constructed in a variety of ways. The most preferable for MOSAID is the running royalty rate that is a percentage of the sale of the chip.
That’s negotiated in a variety of ways, possibly on a country by country basis or using a world wide effective blended rate that provides for the most ease of accounting. It is simply a negotiation. We will have lots of data points on our side. We know what other companies are already paying. Usually it will not be the first negotiation for the licensee. They know what they think they should pay in comparison to other deals they have made. It is primarily a business negotiation.
If you see a company is in possible infringement of your patents, it is likely that they are selling products worldwide. Do you go after them in their home country or in the US?
When you say go after, we are of a mind set that litigation should be the last resort. There are many patent holders who file first and talk second. We prefer to do it the other way around. We approach them at their headquarters whenever possible. That is where the decision makers are. So we will try to communicate directly with the people who can settle this there quickly. But it depends. There is no such thing as a world wide patent. Our patent portfolio is primarily US but we also have patents in Japan, Taiwan, China and different countries throughout Europe. Depending
on what our particular patent coverage is with respect to the patent family we believe has been infringed will determine how we run our assertion program.
If the patent is in the US and there is a company possible infringing on that patent but that company does not directly or indirectly sell products in the US, what recourse do you have?
Because the US is such a huge consumer market particularly with semiconductors a company located in Taiwan, manufacturing chips there on a wafer and packaging them in the Philippines and shipping them to distributors in Hong Kong may argue very strongly that they are not exposed to the US patent system. They are most likely wrong on that point. At least 25% or higher of their end product will end up in the US. Once these products are here, we can attach litigation to those products. That is how we can enforce our US portfolio against a company that places products into the stream of
commerce that ends up in the US.
Theoretically, if an overseas company developed, manufactured and sold a product only in its own country, would you have any recourse with just a US patent?
No! An example of that might be infrastructure equipment for wireless towers. If we had a patent just on that and the towers are physically installed in a geographical location such as mainland China, that’s unlikely to be an export product. It is likely to be sold to be used domestically as long as packaging, manufacturing, design, sales effort and everything else and the physical location of end shipments are in China, the only recourse would be Chinese patents.
Do you agree with the following: The US patent office does not have sufficient expertise quantitatively or qualitatively to cover all the patents that are being applied for. One gets the impression that patents are granted on the theory that they will be later challenged by impacted and knowledgeable firms, if they should not have been granted in the first place.
That’s an over generalization. Certainly, there have been complaints that the quality of patents has declined. Statistics actually the last few years show otherwise. There is a rigorous quality process in the USPTO. There have been problems in the past with employee morale. There has been a high turnover in examiners. Examiners that they have been able to hire have been well qualified. It is a very segmented story. There are very good examiners in very small niches of technology. There are some very thoroughly examined patents that are very high quality out of one examination group and then the quality varies because of different examination groups are unable to attract people with the right background and abilities to do a great job. There is always going to be a difficulty in a large government organization. The patent office is no exception. There are pockets of excellence. They are doing a great job and are well respected throughout the world. The electronics art is one of them. A lot of the criticism of the patent office goes towards other areas such as business methods patents that are going on that are very broad in select areas. In the semiconductor
world there is a huge body of arts and techniques that are used by the examiners. They are well trained and have access to a wide variety of printed publications that help them compare it to the patent applications and to prevent broad claims that are knowingly invalid going through. I guess I am more of a fan than a critic of the current operation of patent applications.
In the case of scientific journals they have peer review rather than rely on the editors as a way to prevent flawed science from appearing. Most of these journals are very specific and it is possible to find qualified reviewers. In the case of the USPTO the breadth of the areas to be covered is almost limitless.
Again, they have narrow assignment of specialization. They will have someone focused on DRAM circuit technology. It won’t be all circuits. It may be cell circuit technology. That’s what they will do all day long. Then they become very knowledgeable in a short period of time. It is not quite as hopeless as it sounds where one guy has to examine combustion engine technology one day, sewing machines the next and then cell phones the following day. They really do have refined practices there.
You touched on business practices where some broad claims have been allowed to seep into the patent system.
I think it is a lightening rod for criticism throughout the world. Business practices are not allowed in many foreign countries.
So you are fairly positive about what the USPTO is doing in the semiconductor area.
Yes. There are enough checks and balances. You can challenge in court. You can also do a re-examination directly with the patent office. There are plenty of ways to improve the quality post issuance. The quality in pre-issuances is now increasing. The vector is going in the right direction.
USPTO. He pushed the two firms to settle where RIM was facing a likely injunction against further manufacture and sale of its products. RIM settled for $612 million. RIM still maintained that the patents should never have been granted and may one day be rescinded.
That’s our legal system. When a dispute is that close, where you have disagreement between a court system and the patent examination bureau, the issue is not black and white. Those are the most difficult cases to settle. What is clogging the court dockets are more of where the infringement is pretty clear and they are just trying to file a litigation to wear down one side or the other to come up with a solution that could be achieved very quickly without the expense of all the legal fees and clogging of the court system. That’s arguably a very good use of the
legal system, when it is a close call and you have to let a third party make the ultimate decision.
In this case the legal systems were not at variance with one another. The patent was granted and the jury trial was held. You can’t argue in the trial that the patent was invalid. So the judgment against was made and they went to the patent office. The process of re-examination took too long so the judge in the civil case had enough.
That could be appealed as well so the court system could decide everything. It is often that invalidity arguments are made in front of judges rather than the PTO. RIM chose a parallel path.
At a high level are the criteria for granting patents similar from country to country?
Many nations have a signed an agreement related to patents. The laws are very similar but the big difference is enforcement which is very spotty in some countries. There is not a fundamental different in terms of the patents or the standards. The enforcement makes the difference.
Is that agreement the Paris Convention (for the Protection of Industrial Property)?
That’s one of them. There is also the WIPO (World Intellectual Property Organization).
Again at a high level are the legal resources of patent holders and the alleged infringers similar from country to country?
Yes, at a high level. The available damages are different in different countries. In many countries they are extremely low. So really what you depend upon is your ability to get an injunction and then force a settlement. It is not even worth the price of going into court to get a full judgment because the damages available are very small. Another thing that differs slightly among different countries is the ability to get injunctive relief.
At a high level they are roughly the same. There are a few variations about whether or not you can even raise as a defense in the same trial whether the patent is invalid. In some jurisdictions you are not allowed to do that. In China, for example, you go back to the patent office to invalidate. They have the exclusive jurisdiction over the right to invalidate. That can kind of screw up a plaintiff. If a patent holder sues somebody in China, they expect to get the trial done in a set amount of time. But it is totally in the defendant’s discretion to hijack the whole
proceeding because now he can start a separate serial proceeding that has to go forth in this state owned intellectual property office, the patent office, for invalidation before we can really go back to this trial on infringement.
It is generally agreed around the world that the US has been for many years the strongest enforcement environment for patents.
Is MOSAID involved in any standard making organizations or efforts?
We have been member of JEDEC for a while.
I assume that you are familiar with the Rambus case and more recently the Qualcomm case. Today do standard making organizations seek to avoid incorporating patented or patent applied for technology into standards without prior agreement on royalties?
Editor: The case against Rambus began in June 2002, when the Federal Trade Commission (FTC) alleged that Rambus had convinced industry groups to declare a standard technology for the memory used in PCs, servers, printers, and cameras without admitting that it owned the patents to those technologies. Among other actions the Commission issued an Order setting the maximum royalty rate that Rambus can collect on sales made of certain JEDEC-compliant parts. Rambus is appealing.
No I do not agree with that. There was certain noise after Rambus. I do not think we have gotten that far as yet. I think that would discourage participation of a lot of companies in the standard process development. I do not think it is a healthy development. What has been common practice to date for patent holders in particular the big companies that you want to be part of this process have a tradition that they will license on a fair, reasonable and non-discriminatory basis. But it depends on getting a grant back or cross license on the same non-discriminatory basis. That leaves room for reality. If you force me to say as an Intel that I am only going to join this standards body with a commitment to license all of my patents for 1% for example to all comers, that leaves you little room with what is actually going to happen. Because another company that is of equal size could come in and say I will take that 1% but I am going to charge you 5% because your exposure is larger, I have a better portfolio, whatever. There has to be some factors taken into account for different values of different cross licenses for example that will always end up in a final agreement and different exposure rates that different companies have and other things that get throw into the mix. If you are going to force the issue to have big companies like Intel sign license agreements only covering patents in an effort to get a standard but they also have 15 other patents that the target company infringes, are you going to prevent them from merging them all into one economical and logical patent license agreement that covers a variety of technologies. How do you compare that to another company that is only exposed to a handful of patents? It is complex. The way it
is now is fine. I do not think there is a need for this drastic change in policies.
It is essentially price fixing which is an antitrust violation. A bunch of competitors sit in a room where you all agree to set a reasonable royalty rate with exact specificity for a patent. I do not think that is a healthy way to go.
believe the patent should not have been granted (unconsidered prior art, overly broad claims) the right to call for re-examination.
That’s a controversial one. What has been proposed in some of these patent reform bills (you are correct that none of them has been past yet) is that there are two windows of opportunity. One window is immediately at issuance. There would be a period of time during which anyone can file an opposition. That’s not as controversial because many other countries already have this. Here is yet another difference between the laws of different countries. In the US you can also ask for a re-examination. You can do that today at any time after issuance, during the pendency of a case or during the life of the patent to the USPTO and claim to have new prior art that wasn’t considered by the original patent examiner and therefore the patent should be re-examined. So this accomplishes essentially the same thing as the opposition procedure. It has not been as critical to get this through but there are a bunch of proponents if this. I think I would be supportive of that first type of opposition myself. The one I am not supportive of is the “second bite of the apple”. A company doesn’t want to monitor every newly issued patent in a particular
area and catches it only whenever that patent holder comes after them which could be years later. That to me doesn’t serve anyone very well. All it would do is postpone negotiations and add expense to the process. They already have the right to sue in court for a declaratory judgment, if there are sufficient allegations of infringement and a controversy. They could go and take their case to court. Or they could use this third party re-examination process. There is no need to create yet another tool in my view.
This week the Supreme Court in a 7-1 decision overturned a ruling that Microsoft should be held liable for patent infringement of AT&T patents on copies of its Windows operating system sold overseas. The court held that software was more closely related to a blueprint than actual components. Do you agree with that decision?
I don’t want to go on the record. I have not read the case yet. I would have to study the decision.
Editor: Jon did agree to provide some backdrop about the case.
The legislation in question was to close an existing loophole. If I have a company that is going to make a ball point pen that comes in two parts, a top and a bottom, that have to be screwed together. And there is a guy that has a patent on the method of making a ball point pen by screwing A into B. The first firm does not want to trip over that patent but it still wants to get the benefit of that technology invention. So they ship the halves across the Canadian border, have somebody screw the parts together and ship them right back across the border thereby avoiding
patent infringement. Congress closed that loophole years ago in Section 271 of the patent code. The Microsoft case is basically the same principal but they are shipping software to Europe for loading into laptop computers claiming that software is not like one half of a pen. It is not hardware and Congress intended to cover only combinations of hardware.
There was a second recent Supreme Court decision around the issue of obviousness. Justice Kennedy wrote "The results of ordinary innovation are not the subject of exclusive rights under the patent laws. Were it otherwise, patents might stifle rather than promote the progress of useful arts."
applying that test and allowing patents that really should be invalidated for obviousness to stand. The Supreme Court dial that test significantly and may have thrown it out all together. It depends on how broadly the case will be interpreted by the Federal Circuit.
Almost every litigation now will be affected where the prior art is based upon a combination of references. They will argue that this case strengthens invalidation arguments.
Editor: The “teaching, suggestion, or motivation” (TSM) test, is a test under which a patent claim is only proved obvious if the prior art, the problem’s nature, or the knowledge of a person having ordinary skill in the art reveals some motivation or suggestion to combine the prior art teachings.
If this test is thrown out, did the Supreme Court suggest an alternative?
They did. The case goes into some detail on what a proper analysis should be with respect to obviousness. The Federal Circuit saw this coming. Once the Supreme Court took this case, they read the tea leaves and saw that the Court was going to review it. It was probably because they don’t like our analysis. They slightly tweaked the test, changed the test in current pending cases. They issued several decisions under a new test. The Supreme Court said they tended to be in agreement with this latest thinking. So it may not be as revolutionary as some commentators are
I just saw that Vonage is going to appeal a recent decision in favor of Verizon on the basis of this decision.
I think there will be some challenges based upon prior judgments having been decided incorrectly. But the affect of that and whether or not the appeals will succeed will be very case specific and very fact specific. If you have a couple of prior arts that you are arguing should have been combined and invalidate a patent, it very much depends on the type of prior art and the context. It is not going to be a blanket, one size fits all affect of this ruling. It will really matter on the specific arguments of each case.
Given MOSAID’s patent portfolio, is there a sense of what the possible impact will be?
We have 750 patents and patent applications. As in any large portfolio, you have a variety of types. Some of them are fundamental breakthroughs. Some of them are at the other end of the spectrum, very minor improvements of existing techno logy. It varies. In all cases, if you were to litigate some of the things that are minor improvements, the defense would be that this case would strengthen any prior art arguments. But the fundamental patents where problems were solved only by the inventor at the time, where it was really breakthrough technology, of which we own quite a few patents,
there should be no effect at all.
Editor: On May 7 MOSAID unveiled HyperLink NAND (HLNAN) Flash, a new Flash memory architecture and interface that will dramatically improve the performance of computer products that use Flash memory, including solid state drives (SSDs), Flash memory cards and USB Flash drives.
The top articles over the last two weeks as determined by the number of readers were:
MAX-View will load and display a chip design with one trillion transistors. Redisplay is real-time on the "Trillion" chip – instantaneous at any zoom level. Micro Magic is providing free MAX-View licenses for a limited time to qualified users. All free MAX-View licenses expire after DAC 45, on June 14, 2008. Retail price of $6,000 per year waived. $2,000 Calibre® interface option also included for free.
AMD Delivers the Ultimate Mobile Platform with New 65nm AMD Turion(TM) 64 X2 Dual-Core Mobile Technology
You can find the full EDACafe event calendar here.
To read more news, click here.
-- Jack Horgan, EDACafe.com Contributing Editor.
Be the first to review this article