Get all 3:


                Learn more

InterDigital...Same but Different

publication date: Jul 3, 2011
Download Print
Rambus and InterDigital are same but different - The Value of R&D and Patents


Rambus is not unlike InterDigital in most respects.  The company develops most of its technology in house and has a high ratio of engineers in the employee head count.  While the attention Rambus receives (perhaps rightfully so due to the perennial nature of the issue) is on unresolved litigation, the business model of Rambus and InterDigital is actually very similar.  Both companies essentially develop and license technology for others to use.  They do not make their own products.  Each enjoys low capital requirements yet works on technologies often well ahead of market adoption.

Not making their own products gives each an additional advantage.  Neither company has to worry about other companies patents being used as countering nor "cross licensing" weapons to the royalties they seek.  Nokia has mouthed an insistence that they should have a lower rate from InterDigital than others (which actually makes some sense from the standpoint of contribution to wireless) but Nokia cannot "attack".

What InterDigital does that Rambus probably wishes it had done is license end products.  That is, instead of licensing specifically the chip manufacturers that produces chips using their technology going into a billion handsets, InterDigital instead licenses the billion handsets on a per unit basis.  By contrast Rambus' licensing in the DRAM industry is aimed squarely at the DRAM chip manufacturers no matter what system or device the DRAM chip winds up being used in.

This end user approach has several advantages but is not without pitfalls.  First there are many more handset makers available as customers than there are DRAM makers.  This is important when it comes to companies that have not taken a license forcing the company to resort to litigation.  On the one hand needing to potentially litigate against or license many parties requires a lot of licensing activity spread about.  However not all the parties in wireless would ever contemplate an expensive litigation battle; so the spread out nature of the licensing model also delivers low hanging fruit. 

By contrast when there are only a few chipmakers as there are in Rambus' DRAM licensing model to go after; any one of them represents a significant enough share of the market to require urgent attention.  In fact the differences in rates of one company versus the other (i.e. one has signed and one has not) is not even considered tolerable because of the already low margins of the chip makers.  There is no low hanging fruit for Rambus (unless you consider a decade of litigation low).  Rambus's fruit is best harvested all at once so parties all have the same expense; but that has been impossible for them to pull off.  Now that Rambus has signed Samsung (who is best positioned to deliver good margins to begin with) to a rate it does not want to see either Micron or Hynix get its arguably even harder to come to an amicable license with Hynix and Micron.  But for the fact that those two parties currently face an antitrust suit and may want out from under it, the litigation would otherwise run the full course without settlement.

Another advantage of the end product licensing approach of InterDigital is that there are several parties using the technology who absolutely know that they are using many wireless technology patents that may (read: do) require licensing.  That is, there are product manufacturers seeking a slice of a large consumer market and have many multiple licenses including those with Qualcomm.  The licensing of this sort of product focused entity is basically to secure outsourced R&D.  Their product interest is not in spending the R&D resources in house to enable them to develop next generation 4 and 5G wireless standards to make their products better.  That's what they pay InterDigital and Qualcomm for.  Thus when approached for wireless licensing these product companies can also be low hanging fruit who recognize the value of the relationship outsourced R&D relationship. 

In reality had Rambus approached DRAM licensing in the end product model (something they briefly toyed with prior to further  developments in patent law in the area of patent exhaustion),  they likely would have a healthy stable of large and small licensees already in systems as well as end products.  The licensing agreements with Intel, AMD and Fujitsu, and Toshiba seem to bear this out.  Instead Rambus has a very few stubborn, large, and "margin challenged" companies that joined together to fight them in litigation. 

Standard and Non-Standard

Lastly, InterDigital (like Qualcomm) participates in the industry standard setting process to a high degree.  Thus their technologies are readily promoted and willingly adopted with full knowledge that a reasonable and non-discriminatory (RAND) license from the company is needed when practicing a particular wireless standard.  This means that companies already know they do not have much (if any) choice about whether or not they will use InterDigital's patents when practicing the standard in their products. That's a good thing when you can extract royalties from 2G, 2.5G, 3G, and 4G/LTE.  It's also the reason we refer to InterDigital as a "baby Qualcomm".

Rambus by contrast has been shaken out of the industry standard setting process because of a bad experience (and unfriendly policy) of the JEDEC DRAM standard setting body.  Antitrust charges were leveled against Rambus for failing to disclose patents and patent applications that they knew would need to be used to practice the standard for DRAM.  While Rambus escaped harm for this in the U.S. even after the FTC Commission (however dubiously) had found fault, the company settled a similar complaint in the European Union with agreed upon lowered royalty rates. 

Fortunately for Rambus the EU settlement only covers technologies in standards that are approaching obsolescence.  More advanced DRAM standards are not saddled with the restrictions on Rambus royalties governed by the EU settlement.   

Rambus now avoids the standard setting process, even though it would make clear the licensing of Rambus IP was required to implement the standard, because the standard setting body takes too stringent a patent position on participants.  That is, Rambus is unwilling to participate when the policy is "give us your IP for free or peanuts".  Instead the policy has become a de facto "we'll just cherry pick your patents with slight "possibly non-infringing changes" and pay you later if you ever win in court".  

In the long run it would make life easier for all if Rambus helped guide standard setting to their IP with all parties aware of rates for the patent licensing before adoption.   JEDEC needs to have a more reasonable policy however for a licensing only company like Rambus not to get fleeced.  The idea that JEDEC is producing DRAM standards free from IP has long been fantasy.  Given the long running feud with Rambus and major players in JEDEC we won't hold our breath on Rambus actively enabling JEDEC standards the way InterDigital and Qualcomm actively enable wireless standards.  But from a long term strategic point of view the relationship Rambus has with its primary industry is simply dysfunctional and highly inefficient for all the parties.    

Thus Rambus has a much higher litigation risk associated with its DRAM licensing and DRAM standards (which is essentially the entire DRAM market) than InterDigital or Qualcomm does with wireless products and standards.    

This is not to say that InterDigital has no litigation risk.  A longstanding feud with Nokia has impacted share price in the past and could again depending on the outcome of an appeal.  In one sense Nokia's dwindling market share is currently welcome news to InterDigital because market share is being taken by companies carrying a royalty bearing license.

We like both companies as technology enablers with a lucrative business model.  While IDCC is arguably more stable because of the complicated litigation issues Rambus remains involved in, each has significant upside and we view each as undervalued for the long term.  In addition, the potential windfall from Rambus' antitrust suit versus Hynix and Micron can produce an explosive gain in shares.

Rambus and InterDigital are same but Different - The IP Model


One of the most attractive aspects of an intellectual property licensing model is the ability to raise revenue without raising expense.  The expression extra revenues "fall straight to the bottom line" can be applicable when the business is running properly.

Consider the DRAM manufacturers.  If they sell $24 billion of DRAM it costs them more money to produce and sell $32 billion.  And heaven forbid they actually need to build another multibillion dollar nanoscale manufacturing plant to get that product.  This sort of high capital outlay and variable cost is just not part of the Rambus IP licensing model.  Each additional dollar of revenue can come with the same assets and licenses in place.  Similarly, InterDigital takes in more money if a licensed company gets a hot phone and sells more in the market.  Not a dime of additional expense is needed to bring in that revenue. 

Of course litigation costs are not trivial in the model.  Rambus in particular has had high litigation expense as a percentage of total revenue for some time.  In fact until Rambus actually secures victories in litigation that reach the top and/or bottom line, the number of cases it has ongoing acts much like a variable cost model.  That is, to gain more potential revenues Rambus has to sue more people.  It has.  Unfortunately this has led to an unshakable market perception about the company as being serial litigators. 

In reality the problem of litigations costs would hamper any smallish company that has yet to fully prove its intellectual property assets and secure more stable revenue streams and profits.  A favorable resolution however can change the fortunes dramatically.  This is why we are actually rather encouraged that in the recent remand of the Hynix suit, Rambus stands in the position of having infringement and patent validity found and need only survive the impacts of spoliation on the case. 

While the antitrust proceedings are certainly just as impactful, the reliance on antitrust damages is not a forward business model.  Still the antitrust case could certainly result in the same impact as proving the value of the IP and securing forward licensing (as well as a nice one time gain) from the small set of licensing parties that would be compelled to end the litigation marathon madness with Rambus.

InterDigital's Recent Boost (IDCC)

Shares of IDCC got a nice 15% boost late this week on news of the final tab of a sale of Nortel Networks patent assets.  The winning and whopping $4.5 billion bid was put in by a consortium of members that includes Apple, Research in Motion, Microsoft, Sony, EMC, and Ericsson. 

Google, who had kicked off the bidding with a starting bid of $900 million, is now left out on what are obviously seen as strategic patents.  That's why it's easy to see why InterDigital would get a bid after the announcement of this mega sale of patents.  One obvious possibility is that Google may still be looking for some patent assets of its own to wield as cross-patenting ammunition if "the consortium" comes calling with infringement against Google's army of Android devices.  

Microsoft notably was already claiming access to Nortel's patents even if they had not participated in an acquisition as they carried a perpetual license.  Could an assault on Android's "open source" development be in the offing?  No matter how you look at what the consortium plans to do with these assets it can certainly be said that the members might largely represent an anti-Android consortium.

What's also interesting is to see Microsoft team up with Apple.  A tag team versus Google involving two of Google's key rivals likely has Google's attention in the area of IP.  Some of this goes to the heart of what we have described as Google's access mission.  They are happy to rinse the profitability out of the handset business making it a commodity vehicle via an open source OS to get users to their bread and butter internet search and ad model.  Apple has come further along this low cost way than we ever thought possible but they currently have the luxury because of several factors.  One is they have an apps market delivering extra profit.  Two they use the brand to wring so much profit out of the marginal increase in NAND when they sell a device with more memory.   NAND is actually providing Apple with price elasticity to get low priced phones into the market as well. 


Nortel Patents: Litigation Readiness or Litigation Avoidance?

In many respects the readiness for and the avoidance of litigation may be the same thing.  Each motivates the consortium players slightly differently.  Apple has recently settled patent issues with both Nokia.  Their motivation for having their own set of cross licensing ammo was already obvious because of suits they were already fighting.  Microsoft, having partnered with Nokia for the OS in their attempt to rebound in smartphones, certainly has an interest in thwarting Android.  Microsoft having additional network related IP that might tie to what Google is claiming as "open source" is also an obvious motivation.

Google is still left in the position of thinking they need patent assets in the avoidance and readiness game.  InterDigital may not actually be the most likely candidate for aiding Google due to the high prevalence of agreed upon RAND licensing in their wireless standard activity.  However the sheer size of the value of the Nortel patent assets in the recent sale does open up the eyes of the market to the value of InterDigital's patent assets and licensing model.  Google could be interested.

What may make InterDigital attractive to Google however is what we discussed above.  In the Rambus and InterDigital model each company acts as an R&D house for their industry.  In Rambus's case that industry has thus far chosen to use the R&D but only pay if forced.  But there has to be a running joke at JEDEC every time a Rambus innovation is introduced as a "free from patents" addition to the standard.  This running joke makes Rambus frustrating.  When viewed from the standpoint of being paid for its technical contributions Rambus is indeed frustrating.  On the other hand the same running joke makes it a "can't miss licensing all DRAM" proposition at some point.   The JEDEC process is simply accustomed to using the Rambus IP at this stage.  If companies eventually pay they seem "no worse off" so it won't stop anytime soon.

InterDigital, like Rambus, has always worked to be a generation ahead.  Their work in 4G for example was heating up while the world was pondering "when will 3G" really take off.  So from the prospective of Google you might get two things with InterDigital.  The first is a de facto R&D arm that is already an essential element of wireless standard development.  The second is your own wireless R&D group armed with patents not yet incorporated into the next generation of wireless.  So, as of the multibillion dollar Nortel IP sale, we would have to agree that the acquisition of InterDigital by Google actually would make some strategic sense especially if they want to ever make their own hardware.  The difficulty would be determining how to continue to ring the royalties out of licensing IDCC's assets when the subject is likely to turn to cross licensing with consortium members.  But if Google continues in the direction of not making its own handsets the ability to avoid cross licensing can easily remain intact with the real exposure to Android coming from software and network software IP in the hands of Microsoft and the consortium.

We remain bullish on InterDigital.