Tuesday, September 6, 2011

ITC review of S3 Graphics v. Apple: things may get better or worse for Macs, iOS devices seem safe

On Friday afternoon, the ITC published its notice of a Commission determination to review the initial determination made by the Administrative Law Judge investigating S3 Graphics' patent infringement complaint against Apple.

What lends that case strategic importance is the fact that S3G is in the process of being acquired by HTC, which is under serious pressure after another initial determination found its Android-based devices to infringe two Apple patents. HTC is trying to get that one reversed, and I recently discussed the related petitions for review and discovered, in a fairly recent Apple pleading, a passage that claims Android founder Andy Rubin got the inspiration for the Android framework while working for Apple.

As I wrote in late July, the S3G action is, relatively speaking, HTC's best near-term chance to have serious leverage against Apple, which HTC needs badly in case Apple's first ITC complaint succeeds since Apple isn't too much into outbound licensing and won't just do a deal on a $5-per-unit kind of basis, but based on information that became available at the time, the S3G case looked "increasingly unlikely to help HTC against Apple". I said that the largest part of Apple's U.S. revenue base -- the iOS devices (iPhone, iPad, iPod) wasn't held by the ALJ to infringe any of S3G's four asserted patents, and it looked like Apple could easily solve its Macintosh-related patent infringement problem by purchasing NVIDIA graphics hardware (though this would likely require some engineering changes).

Having just spent two days analyzing the ALJ's initial determination in light of the questions raised by the Commission (the ITC's highest-ranking decision-making body) in the notice to which I linked in the first paragraph of this post, I have to modify my assessment -- not with respect to Apple's iOS devices, which appear on the very safe side for now, but as far as the NVIDIA solution for Apple's Macintosh computers is concerned. For the Macs, there's a significant risk of things getting worse, but Apple could also achieve an improvement by having all of the asserted patent claims declared invalid. Anything is possible now since the Commission decided to review the matter "in its entirety", and either party will see opportunities in the questions that were raised.

Further below I'll outline the results of my in-depth analysis (of hundreds of pages of legal documents) in more detail.

Potential Mac/mobile conflict of interest on the horizon

There's a certain possibility that Apple might have to decide in a few months whether it keeps up the pressure on HTC with a view to mobile devices and refuses to settle even if there might be a substantial opportunity cost for its Macintosh business in the U.S. at least until Apple perhaps succeeds in having the relevant claims invalidated (which is a possibility since a declaratory judgment action in California as well as reexaminations by the US Patent & Trademark Office are underway, and on a purely preliminary basis, all of the relevant patent claims have been found invalid) or until HTC will be forced to settle everything on Apple's terms.

By "substantial opportunity cost" I don't necessarily mean lost sales and market share due to an exclusion order: Apple could also decide to switch to an alternative, non-infringing, graphics system or alternatively have at least the final assembly of its Macintosh computers done in the United States, which (depending on what exactly is imported and whether the required manufacturing resources can be identified in the United States) might be enough to work around an import ban and would be more viable for Macintosh-size computers and notebooks than it is for smartphones and tablets.

Assembly in the United States would reduce Apple's margins in that business area, but that's the price Apple might have to pay -- and might be quite willing to pay -- in order to maintain its full leverage against HTC for some more time. Compared to the threat Android represents to Apple and the cost of its legal battles, the higher cost of a final assembly of its computers in the United States might be just another investment in the ongoing intellectual property war. It could be a rather temporary opportunity cost since Apple might get those patents invalidated anyway, and in that case an import ban would no longer apply. At the same time, Apple might make so much headway against HTC's Android-based devices that we might only be about 12 to 18 months away from a point at which HTC would have to settle on Apple's terms anyway. In that case, Apple could also solve any problems S3 Graphics might (or might not) pose to its Macintosh business.

In an intellectual property war, a party's financial strength -- not only in terms of money in the bank but also in terms of the ability to handle margin hits (especially if likely temporary) and a diverse revenue base -- can be a decisive asset against a substantially weaker rival.

HTC and the Fata Mo[to]rgana

I like portmanteau words and am now going to present a new one:

Fata Motorgana = "Fata Morgana" + Motorola

Like a Fata Morgana that leads people to see castles in the air, a Fata Motorgana is also a mirage. It's the fallacious assumption that Google's proposed acquisition of Motorola Mobility (MMI) is going to help Android OEMs such as HTC with their patent worries. The truth is that those patents aren't going to give Google the necessary leverage to protect the Android ecosystem at large. Worse than that, not only is "Googlorola" not a solution to those problems but it will require all other Android device makers than MMI to compete with the maker of their platform, a game that they're never going to win.

Officially, HTC welcomed the Googlorola announcement. But what else should HTC's management do? Since the first half of July, HTC's stock price has been under serious pressure with what is clearly a downbound trend over the last few months, despite a stock purchasing program and other attempts to prevent the worst. If HTC's management spelled it out and said "Googlorola isn't going to help us against Apple and instead they're going to undercut and outperform us in the Android market", there would be panic. They obviously don't want that. So they have to put on a happy face even though they may already be thinking very hard about alternatives to Android. (Apple just pointed out to the ITC that HTC also does Windows Phone devices, which aren't affected by Apple's litigation.)

The Fata Motorgana phenomenon has already resulted -- and will continue to result -- in claims by certain analysts that are sometimes just off-base. In fact, only a minority of analysts got it right early on. Most spoke out on this without even understanding the patent litigation and licensing business well enough. There was even the suggestion that MMI was going to help Samsung. However, while the average per-patent strength between Samsung and MMI is at a similar level, Samsung has about 100,000 patents worldwide, roughly five times more than MMI. In fact, Samsung has far more patents in the U.S. alone than MMI has patents plus pending applications worldwide.

But it's a fact that one can fool some people for all time, all people for some time, but not all people for all time, and that's why more and more people are figuring this out. It was a wake-up call to some of those "mutually assured destruction" dreamers when M-CAM founder Dr. David Martin said on Bloomberg TV that MMI's patents are "crap". Nokia's patent chief was quoted by Reuters as saying that the MMI "won't solve one" of the dozens of ongoing Android-related infringement lawsuits. Not even one. I agree.

So if it's not really a patent protection kind of deal, what is it about? It's all about control and locking in end users. Concerning the chances of the likes of HTC to compete with a Google-owned MMI, I strongly recommend this article by bnet's Erik Sherman on an interview in which Google's executive chairman and until-recently-CEO Eric Schmidt said that the deal was about much more than patents. Two spot-on quotes from that bnet article:

  • "Smartphone makers now know that Google plans to take them on directly, no matter what it says going forward."

  • "Who wants to go to all that [patent] trouble when Google also wants to kill your hardware sales in favor of its own?"

HTC's purported fear of critical voices

I just talked about how those patent problems (and probably also investors' increasing understanding of Google's true agenda for MMI) affect HTC's stock price. In this context I'd like to mention -- as a matter of transparency -- that the Wall Street Journal recently wrote that conference calls organized for me with clients of investment banks Jefferies & Co. and Macquarie Equities Research "were canceled after HTC complained sources familiar with the matter say".

The WSJ says HTC declined to comment. So did I: all that I can confirm is that these brokers had indeed sent out invitations to their clients for conference calls with me (in July), shortly after which the calls were canceled. I don't know what was going on behind the scenes, and I don't know who the WSJ's sources are (it's a reputable paper that's extremely well-connected in the financial services industry). However, I can say that I would have been available for those calls as planned, and I continue to hold those two banks in the highest regard.

It's also no secret that I advise professional investors all the time, and when I do so, I do my best to deliver accurate representations of the intellectual property situation and how I see it from my vantage point.

I'm always receptive to information from industry players. I have contacted, as well as been contacted by, the PR departments of several of the world's leading mobile device makers and other relevant companies, most of whom are key players in the Android ecosystem. None of them may have ultimately agreed with my reporting and analysis in full, but I do want to engage in constructive dialog with all those who are interested in it.

Apart from the fact that I don't know whether it's true that HTC played a role in the cancellation of those conference calls, it wouldn't affect my view of HTC anyway. Also, while its acquisition of S3 Graphics won't be the definitive solution for its patent problems, the deal might still have made business sense (at least more sense than the idea that Google offered $12.5 billion primarily for MMI's patents).

My analysis is predominantly based on what I see in court (including ITC) filings. Looking at the hard facts, I have arrived at the conclusion I outlined further above, and in the remainder of this post, I'll explain in more detail what the key issues in this ITC investigation -- and their potential implications -- are at this stage of the proceedings.

The key issues in the Commission review

The ALJ's initial determination in such a case is key, but if and when the Commission decides to conduct a full review, it is not bound in any legal way by what the ALJ wrote. It can decide de novo, as if starting from scratch, though that rarely happens. Also, no new evidence can be presented at this stage.

The review notice raises thirteen questions on which the Commmission would like the parties to comment. While those questions don't formally limit the scope of the review, they provide an indication as to which kinds of issues the Commission considers particularly important and of which it isn't necessarily certain that the ALJ reached the correct conclusion.

The last one of those questions relates to a possible bond (if the ITC orders an import ban). That's a routine question of no relevance to my analysis of the possible outcome.

The other 12 questions touch on three kinds of issues:

  • The first five questions are of a jurisdictional nature. They relate to the scope of the ITC's power in cases in which a product isn't necessarily infringing at the time it is imported into the United States but is subsequently used in an infringing way. In those gray-area cases, there comes a point at which a right holder cannot expect the ITC to help but has to go through the (usually slower) federal courts.

    This first group of questions appears to be the most important one to the ITC, which is not surprising given that this is an institutional issue. On the one hand, the ITC can't overstep the boundaries of its jurisdictional power. On the other hand, if it's too easy for infringers to import non-infringing material that is, however, reasonably closely connected to an infringement occurring post-importation in the United States, then the ITC would no longer be able to play an important role as an IP enforcer at least with respect to a variety of technology products.

  • The second group consists of four questions related to the concept of patent exhaustion in connection with Apple's use of NVIDIA graphics chips in some of its Macintosh products. Patent exhaustion is the concept of an "infringer" not being liable for using properly licensed components, but whether it applies depends on the specifics of the license and of the infringement pattern. Those are questions #6, #9, #10, and #11.

  • The third group comprises three questions (#7, #8, and #12) related to the possible invalidity of the asserted patent claims. The relevant considerations are connected in one way or another with some of the prior art identified.

Interestingly, there are no questions related to the infringement analysis per se, and not even any claim construction questions that suggest a possible modification of the ALJ's interpretation of some of the key terms (which could affect the validity as well as the infringement assessment). In other ITC reviews, such as the ongoing one concerning Eastman Kodak's complaint against Apple and RIM, claim construction is usually key and often the most important part. There is a question (#5) that looks like an infringement question at first sight but it's actually part of the "jurisdiction" consideration.

Each of the following four sections relates to one category of issues -- starting with the "infringement" question, which the review notice doesn't raise directly, and then the three groups of questions raised by the ITC -- and contains my take on what is more and what is less likely to happen in the months ahead with respect to each category.

Please take two things into consideration with respect to my analysis: I didn't have access to all of the material (some documents aren't publicly available and others are more or less heavily redacted), and I looked at the ALJ's initial determination only in light of what it means in combination with the review notice (as opposed to looking into each and every technical detail myself).

Infringement: iOS devices not likely to be affected

The ALJ's initial determination found all four of the asserted patents infringed by Macintosh computers but only two patents had claims that were deemed both valid and infringed (and only an infringement of a valid patent claim matters).

However, the ALJ didn't find iOS devices and the iOS Software Development Kit (SDK) to infringe any of the asserted patent claims.

A patent claim typically consist of multiple elements (also called "limitations"), and only a technology that fulfills all of those criteria at the same time is held to infringe. It's a logical AND combination of multiple criteria.

The ALJ had several reasons for each of the asserted patent claims why it wasn't infringed by iOS. The Commission would have to disagree with the ALJ's findings in many ways in order to find any iOS product (or the SDK) to infringe any of those patent claims. It could happen, but it seems very unlikely, especially since the Commission didn't even raise any question that reflects an interest in this.

Nor did the Commission ask any question that indicates doubts about the infringement of the asserted claims by Mac OS X from a purely technical point of view. The aforementioned three groups of questions provide opportunities for Apple to make headway and render some of those technical infringements legally inconsequential, but it looks like a long shot to convince the Commission that what those Macs do is technically different from what the asserted patent claims describe.

Jurisdiction: nexus between importation and infringement

I said before that those jurisdictional questions -- the largest of the three groups of questions -- are particularly important to the Commission.

They are also an opportunity for Apple to avoid an import ban. Based on my analysis, it wouldn't be surprising if jurisdictional considerations took care of any problems Apple might otherwise have with claim 11 of the '978 patent (the only claim of that patent that the ALJ deemed both valid and infringed). The Commission wants to know more about how Apple allegedly infringes that patent claim when it "sells applications containing compressed DXT texture". In my view, the sale of apps -- which occurs after a product has been imported and sold -- may very well be beyond the reach of the ITC. There may not be enough of a nexus between importation and infringement when we're talking about an aftermarket. The Commission might also be able to determine that there isn't enough evidence in the record (and now it's too late to introduce new evidence) to identify an infringement of that kind. This way, the Commission wouldn't have to settle the "nexus" question.

I find it hard to see how the jurisdictional question could help Apple avoid an import ban based on another patent. Should Apple get rid of one infringement issue but still have to deal with another one, it might be in essentially the same short-term position (possible import ban). However, I'm sure Apple will fight for every bit of progress. If some patents are valid but not infringed and others are infringed but invalid, Apple is fine. This could also happen "after the fact" in terms of the ITC perhaps ordering an import ban based on a patent that the USPTO or a federal court might invalidate shortly thereafter.

Also, if there is an import ban but the ITC defines the limits, it might (even if presumably unintentionally in that case) give Apple guidance for a workaround.

Patent exhaustion and the NVIDIA deal

In connection with Lodsys, Apple makes a patent exhaustion argument, claiming that its app developers can't be sued for using an API that Apple -- which is licensed to use the asserted patents -- provides. I already pointed out in that context that patent exhaustion has limits. Whether a license extends to the downstream depends on the specifics of the license and of the infringement pattern.

Apple also raised patent exhaustion as a defense in this case. The ALJ disagreed with respect to Intel. Intel has a license to those patents (which S3G apparently admitted), but the ALJ concluded that "it is software, and hardware components other than the Intel CPU or GPU, that constitute the infringing features of Apple's products".

The ALJ did, however, agree with Apple that its use of NVIDIA GeForce graphics adapters is a scenario of patent exhaustion and, therefore, a "complete defense" for those Macintosh products that include an NVIDIA GPU.

The review notice raises questions about both NVIDIA and Intel. A couple of the questions in this category look non-judgmental, but question #9 doesn't bode well for Apple's purposes. It asks "whether the authorized purchase of a patented component gives the purchaser the right to (a) use its own independent implementation of the patented technology, and (b) the right to use the purchased component in conjunction with other components that together utilize the patented technology". The (a) part appears to be almost a rhetorical question: the answer is certainly "no". For (b), I don't want to commit to "no" being the answer in 100% of those scenarios, but many (and probably most) of the use cases it relates to would also have to be considered infringing.

I think there's considerable risk now that Apple's NVIDIA-related exhaustion defense may fail. It's not over yet, but it looks difficult, and frankly, I can't say that the ALJ's exhaustion reasoning in the initial determination struck me as particularly convincing. He talks a lot about the history of those licenses (the NVIDIA license resulted from a settlement of litigation in California), but I couldn't find a clear explanation of why Apple could implement algorithms patented by S3G in its own software and then use the presence of a licensed graphics chip in the same machine as a defense.

This block of questions constitutes a big-time opportunity for S3G, but again, only with respect to Macintosh computers, not iOS devices (since those don't seem to infringe).

Invalidity: anything's possible

The possible ramifications of the invalidity-related questions raised in the review notice are sufficiently broad that this could go in all sorts of directions.

The Commission might uphold the ALJ's findings: he found all asserted claims of two of the patents invalid (for obviousness) and reduced the other two patents to one claim and, respectively two claims.

Question #7 doesn't provide an opportunity for improvement but constitutes a risk of exacerbation from Apple's point of view (but again only with respect to Macintosh computers, not iOS devices). It's about prior art -- the "Hoffert" patent (U.S. Patent No. 5,046,119, which doesn't explicitly state headers as a criterion, but the ALJ found that it was obvious someone would have used it with headers even at the time when S3G filed for its asserted patents. Examples of headers being used in connection with compressed image data at or before that time include the CCITT fax compression standard, the CGI format, and certain file formats on the Commodore Amiga.

I believe Apple is fairly likely to defend the ALJ's obviousness take concerning headers and the Hoffert patent. But if not, the number of valid patent claims infringed by Apple would likely increase.

Question #8 is about the combination of the Hoffert patent with a writing by an author named Knittel. The ALJ sided with Apple on this one and invalidated a couple of claims on that basis. Compared to the initial determination, Apple can only lose here, but it may very well be able to defend the ALJ's take.

Question #12, however, is a huge opportunity for Apple. It deals with an invention that the ALJ deemed abandoned, suppressed, or concealed and therefore didn't admit as prior art. If the Commission sided with Apple and not the ALJ on this one, Apple would be able to do away with all three patent claims the ALJ deemed valid and infringed.

The outcome of all relevant patent claims being invalidated wouldn't be too surprising. All four patents-in-suit are being reexamined by the USPTO, and first Office actions have been issued, declaring all of the claims asserted in this case invalid on a preliminary basis. You can read more about that on the LitigatingApple blog. You can find the two first Office actions related to the two patents deemed valid and infringed on Scribd. The patent examiner clearly agreed with Apple's invalidity arguments to a large extent, but first Office actions are by definition non-final and final decisions often less favorable to the challenger's interests. Caution is recommended at this stage. Also, the ITC didn't even admit first Office actions as evidence before the initial determination and won't do so now, although it may take note of those findings and be influenced by them at any rate. Or even without taking note, the Commission might also arrive at an independent conclusion that those patents related to inventions that were obvious over the prior art that existed at the relevant time.

With so many open questions, it's not even clear whether the Commission will take the final decision at the end of this review. It might also remand some of the issues to the ALJ...

If you'd like to be updated on the smartphone patent disputes and other intellectual property matters I cover, please subscribe to my RSS feed (in the right-hand column) and/or follow me on Twitter @FOSSpatents and Google+.

Share with other professionals via LinkedIn: