Wednesday, December 26, 2018

List of complainants over Qualcomm's conduct keeps growing as LG Electronics joins Korean antitrust action

While things slow down in Western countries between Christmas and New Year's Day, "Tis the season" has a different meaning in South Korea. Two years ago at this time of the year, the Korea Fair Trade Commission (KFTC) slammed Qualcomm with a fine. And now BusinessKorea reports that LG Electronics has joined the related antitrust lawsuit.

LG's decision is important since Samsung withdrew from the case last year following a comprehensive business agreement with Qualcomm involving patents, chipsets, and manufacturing. However, depending on how the FTC v. Qualcomm trial (scheduled to start on Friday, January 4) will go, Qualcomm may have to renegotiate its deal with Samsung anyway.

According to BusinessKorea, other complainants in Korea include Apple, Intel, MediaTek, and Huawei. MediaTek is mentioned from time to time in connection with Qualcomm's conduct, and one of my New Year's resolutions for this blog will be to find out at least a little bit more about the role they play. The fact that the name of this Taiwanese semiconductor company keeps coming up got me curious.

Meanwhile in Germany, where Qualcomm capitalized on the terrible "defendant's dilemma" laid out by the court, we may still have to wait a little before we know about a potential stay of enforcement of the injunction. Today is yet another public holiday in Germany (Second Christmas Day). Since the ruling came down on Thursday afternoon, the earliest time when Apple could realistically have filed an appeal and a motion to stay enforcement would have been Friday, December 21 (if they had already prepared an appellate brief in anticipation of what happened). But this is a high-profile case, with major economic implications, so maybe the appellate judges are already looking into this despite the Holiday Season. German judges are free to take case files home (it's all on paper though some courts, such as the Federal Patent Court, internally work with electronic documents to an increasing extent) and work whenever and wherever they please. The most fundamental question will be whether the appeals court affirms or reverses the lower court's agnosticism (the court stated in public that there may not even be an infringement).

It doesn't look like enforcement has already begun. Qualcomm firstly needs to post the bond or make the deposit (north of $1.5 billion in the aggregate of two injunctions, a fact missed by fake news media and clueless commentators) and on that basis serve the injunction on Apple. They said it would take a few days, so maybe enforcement will begin later this week.

Share with other professionals via LinkedIn:


Monday, December 24, 2018

The patent defendant's dilemma in Munich: damned if you do, damned if you don't give up your secrets

The shocking part about last Thursday's Qualcomm v. Apple patent injunction in Munich is not that Qualcomm finally won a case after several trials at which the courts in Mannheim and Munich were rather unconvinced of a valid patent being infringed. Much to the contrary, an analyst asked me a week ago: "How come Qualcomm, with its more than 100,000 patents, hasn't just been able to assert a few strong hardware patents and end the dispute with Apple that way?" I told him that chipset patents, while harder to work around, appear to face many of the same quality issues as software patents.

It was to be expected that Quinn Emanuel was going to win something at some point.

The economic implications don't shock me either. All that Qualcomm has to show now, for its north of 100K patents, is an agnostic judgment (the court stressed even in a press release that there may not even be an infringement) affecting older iPhones, and analysts say Apple generates only about 3% of its sales in Germany (which in turn is a significant but far from a huge market for Apple) through its own online shop and the 15 official German Apple Stores.

On U.S. TV I saw an interview where someone noted that the court wouldn't have required a massive deposit if the case wasn't economically significant. In fact, the court had reduced by 60% the amount Apple originally demanded, but that was because it turned out that the alleged infringement certainly didn't apply to the 2018 iPhone models. Still, yes, the amount is significant--but that's in no small part because the court has to assume that if Apple's Ireland-based European distribution company stops supplying iPhones 7 and 8 to German customers, those sales will be lost, and can't just assume that those customers (major carriers and retailers) will simply source elsewhere in the EU's Single Market.

As for the deposit and the old/new iPhone distinction, I'd like to debunk, once and for all, just two of the pieces of fake news spread by woefully incompetent commentators on TV and on the Internet who pretend to know what they're talking about but are just clueless:

  1. The bond or (in the alternative) deposit required under the court ruling is € 668.4 million per case. There are two cases (one against Apple Inc., the U.S. parent company, and another one against Apple's European distribution company and the German entity operating 15 Apple Stores in the country). So the total (since it wouldn't make sense for Qualcomm to enforce one of the rulings but not the other) is € 1.337 billion, or more than US$1.5 billion.

    It's obviously not against the law to comment on a case without at least reading the court's press release, but it discredits the ones who fail to do their job. The court's press release says "Sicherheit in Höhe von jeweils 668,4 Mio. €," which translates as "security amounting to € 668.4 million, respectively," with "respectively" meaning "per case" in this context. The court's press release explains clearly it's about two cases and stated the two case numbers.

  2. At least one commentator on TV speculated that Qualcomm might be able to seek an "extension" of the injunction to newer iPhones and drew an analogy to China. Completely wrong. Under German law (and it's the same in the U.S. unless your injunction is suboptimally-worded) an injunction affects all violations that have an "identical core" ("kerngleicher Verstoß"). So even if the court didn't evaluate newer iPhones, the injunction would apply if the violation was technically of the same nature. But Presiding Judge Dr. Matthias Zigann stated at Thursday's announcement that the reason for which the security amount isn't even higher is because there was apparently consensus at trial that newer iPhones are technically different enough so as not to fall under that injunction.

Even all the fake news isn't shocking because it's normal in connection with high-stakes commercial litigation, and especially high-stakes patent litigation.

No, the absolutely terrible thing that I'm still profoundly concerned about is that Judge Dr. Zigann laid out a doctrine under which Germany would become the perfect patent jurisdiction for companies to steal competitors' technical secrets.

First, Judge Dr. Zigann explained that German civil litigation is centered around the parties' proffers of evidence as opposed to U.S.-style pretrial discovery. The complainant brings an allegation; the defendant must counter it with a denial; and the more substantiated the allegation is, the more substantiated the denial must be. It's a ping-pong game of mutually upping the ante.

In these two Qualcomm v. Apple cases over the envelope tracker patent, the court concluded that Apple's definitive denial (not just a denial for lack of sufficient information to form an opinion or belief) fell short of Qualcomm's efforts to establish infringement by means of a teardown report.

Many media reports quoted Judge Dr. Zigann's "defendant's dilemma" doctrine, and as an earwitness I can confirm that this is what he indeed said:

"If a defense is only possible by means of revealing a secret, a party must reveal it, in which event it ceases to be a secret, or a party may elect not to reveal it, in which event it may lose the case, as has happened today."

Apparently it never crossed his mind that the lesser evil would be to just throw out some patent infringement complaints for evidentiary failure, especially when a plaintiff actually has the evidence and withholds it. The fact that Qualcomm's German counsel elected not to sign the protective order and deny Qualcomm's engineers access to Qorvo's technical secrets should have given the court pause. Why would Qualcomm firstly conduct discovery in the U.S. for the stated purpose of presenting chipset schematics in Munich but then elect not to do so?

It happens again and again, also in connection with antitrust defenses (which are an affirmative defense in Germany only in connection with injunctive relief), that, whenever there is doubt, cases are resolved in favor of patent holders.

We'll see what the appeals court ("Oberlandesgericht München" = "Munich Higher Regional Court") makes out of this. It would be disastrous if it affirmed the defendant's dilemma doctrine because it would create a whole new incentive for patent holders to sue competitors in Germany.

Qualcomm's insistence that the chipset schematics be made available to its own engineers or not be considered by the court and the court-appointed expert witness was unreasonable and should have weighed strongly against the credibility of its infringement accusation, including its inherently-unreliable teardown report.

Share with other professionals via LinkedIn:


Friday, December 21, 2018

Shareholder value analysis: what Qualcomm stands to gain or lose from $1.5B German iPhone ban

Yesterday's final (though appealable) judgment by the Munich I Regional Court--granting Qualcomm a Germany-wide injunction against iPhones older than the 2018 models over a power-efficient enveloper tracker chip design--is not merely wallet paper to decorate Qualcomm's or Quinn Emanuel's offices. Qualcomm immediately announced that posting the bond or making deposit required by the court for the right to enforce the injunction during an appeal (something that Reuters initially got wrong) "will be completed within a few days." Apple, predictably, told the media that it was going to appeal. Otherwise the bond or deposit wouldn't be necessary, but no one could have doubts for even one second whether Apple would appeal.

There are technically two cases (one in which Apple Inc. is the defendant and another one where two European Apple entities, including the German company operating 15 local Apple Stores, are targeted). In the aggregate of the two cases, Qualcomm is now going to post a bond or make a deposit of more than $1.5 billion. As Judge Dr. Matthias Zigann explained yesterday, the amount is 60% less than what Apple's counsel had demanded, and the court felt the massive reduction was justified by the fact that the 2018 iPhone models aren't going to be affected (only iPhone 7, iPhone 7 Plus, iPhone 8, iPhone 8 Plus, and--but it wasn't on sale anymore anyway--the iPhone X). Qualcomm had disagreed with the amount, but failed to present facts that would have persuaded the court to slash it further.

The liquidity cost to Qualcomm will be significant, as an appeal could take about a year to be resolved: we're talking about tens of millions. Even though Qualcomm's present litigation costs are huge, as are PR expenditures that according to the New York Times even included a fake "Draft Tim Cook 2020" website, that's still a lot. But from the perspective of Qualcomm shareholders, anything that would significantly increase the chances of settling the dispute with Apple--plus, ideally, from Qualcomm's perspective, getting rid of baseband chipset competition from Intel--may seem worth a lot more than this. However, the court didn't require $1.5B only to cause a liquidity cost of maybe $30-$50 million. It's because there is, just like there would be in the same situation in a U.S. litigation, a clear and present danger of staggering wrongful-enforcement damages.

Let's look--from a Qualcomm shareholder value perspective--at the costs and benefits, risks and opportunities. After an itemized analysis, we'll get to whether it's worth it.

Please note that the appeals court can stay enforcement of the injunction for the duration of the appeal (and, at the end of the appeal, may overturn the injunction as a whole). The analysis here is, therefore, based on a pure hypothetical: on the assumption, for the sake of assessing risks and opportunities, that the appeals court would not stay enforcement, despite Apple presumably attacking the lower court's ruling from multiple angles now.

Potential benefit #1: immediate leverage over Apple

Apple has already announced that older iPhones would now be removed from the 15 German Apple Stores, though customers will still be able to buy them in other stores and from carriers (for the time being).

Based on what I've read on different websites, the prices of the newer iPhones have recently come down a bit. Without a carrier contract, the 4.7" iPhone 7 is listed on Apple's German website at a price starting at €519 (7 Plus: €649). The iPhone 8 starts at €679 (4.7" display) and €789 (5.5" Plus size). The 2018 iPhone XR, with a 6.1" display (a larger net screen size than the old "Plus" format though the device as a whole is smaller), starts at €849. Looking at those prices, I believe the iPhone 8 was just available for consistency's sake, but if anyone had told me they wanted to buy one now, I'd have urged them to get an XR because the incremental value far, far exceeds the incremental price, making a purchase of a new iPhone 8, at this point, an illogical decision. However, if someone had told me they wanted an iPhone, and it should be the cheapest one, but they wouldn't want a used one, then I'd still have recommended the iPhone 7. In fact, I'm still using a 7 Plus, and a couple of months ago I had its screen replaced at a cost of about €200, which I thought was a smart choice for potentially skipping one product generation.

Obviously, Apple can always reduce the entry point in Germany by just lowering the price of the smallest iPhone XR. But presumably some price-sensitive buyers who want a contract-free new phone for about €500 would have to go elsewhere, or wait.

But it's not just about what the 15 German Apple Stores do immediately. The analysis must be more comprehensive.

The court ordered, besides an injunction against new sales, a recall from resellers (carriers, retailers) and the destruction of all products returned by resellers. With the Holiday Season and everything, it may take a couple of weeks before a letter from Apple to resellers goes out and urges them to return those products. Small resellers may be impressionable; the likes of T-Mobile are themselves exposed to patent infringement litigation all the time, so I wouldn't be surprised should their participation in the recall effort be, well, characterized by a lack of enthusiasm and speed, or even an outright lack of cooperation. At least that's what other patent holders (such as IPCom) alleged in the past. The injunction binds three Apple entities, but no third parties. Theoretically, those third parties must and do know that a willful patent infringement in Germany can even be deemed a criminal offense, but again, those are deep-pocketed, sophisticated organizations and not that easily scared.

In fact, those larger resellers are so deep-pocketed and sophisticated that, if they needed any new supply of iPhone 7 or iPhone 8 units, they could just buy the thing from Austria (not sure whether this works for carrier-specific phones, but certainly for unlocked ones). The German injunction doesn't apply to Austria (a bordering EU member state with the same official language, German), nor does it involve customs seizures (which Qualcomm would have to seek separately).

If Qualcomm sought to accelerate the recall at all costs, it would have to seek injunctions against the resellers. That would position Qualcomm as a patent troll, and would make Qualcomm some additional enemies. And in the case of that envelope tracker patent, Qualcomm would face a high risk of losing any subsequent case, given that the technical win over Apple was not based on conclusive evidence of infringement but on the sequence of pleadings (we'll get to that again in the public relations context).

This is now totally speculative (I can guarantee that no direct or indirect source of any kind said or even hinted at anything like it), but what if Apple actually had the option, at some point or maybe at any time of its choosing, to put all the technical facts on a court's table, and if those facts would then be more or less certain to result in a finding of non-infringement just like at the ITC? Maybe it's just that as long as the damage is minor, Qorvo's secrets will be protected, but at some point, the damage might exceed the value of those secrets? The problem for Qualcomm is that until all appeals are exhausted, those facts can still be put on the table and result in a wrongful-enforcement damages claim against Qualcomm. If Apple's calculus was (again, note what I just said about how totally speculative it is!) that they're going to win this sooner or later anyway, then there wouldn't be that much leverage: while lost sales are always a bit hard to prove (especially in a case where you still have multiple iPhone models on sale and where third parties may still offer them for quite some time), the costs of a recall and of the destruction of such devices can be accounted for, and that could cost Qualcomm dearly in the end.

So the leverage Qualcomm gets from the Munich injunction isn't huge (new iPhones are in most demand and are most profitable). At best, it's "significant but limited," but if Apple can simply play a joker anytime by presenting Qorvo's technical secrets, then it's extremely limited, with the possibility of this ultimately even becoming a net negative for Qualcomm due to wrongful-enforcement damages.

The possibility of Apple actually having a joker up its sleeve is just due to the fact that yesterday's decision doesn't have the kind of technical basis--a definitive identification of the presence of each and every claim limitation in the accused embodiment(s)--that a solid patent infringement finding would need, and even "solid" ones are often overturned. If Qualcomm had asked me whether it's a good idea to enforce, I'd have said: guys, if you had a real infringement finding here instead of an agnostic ruling (where the court itself said they don't know whether there really is an infringement), then, yes, I'd probably understand why you'd take your chances and enforce. But be realistic: the appeals court may simply consider the lower court's agnosticism reversible error (for instance, because of Qualcomm's German lawyers simply not having been cooperative regarding the chipset schematics the same law firm had obtained in the U.S. for the explicit purpose of presenting them in Germany), or Apple may have a joker and it may do away with agnosticism--and then your case here will go up in flames, as will a fair amount of money. And by the way, the deposit is not an upper limit: it's just what the court thought might be at stake, but ultimately you'll be liable for actual damages, which can be more (or, obviously, also less).

Potential benefit #2: war psychology

Even if the questionable leverage (see the previous section) alone didn't justify this risk exposure and the liquidity costs, Qualcomm would likely be prepared--and, honestly, well-advised--to pay a certain premium for displaying a take-no-prisoners attitude to Apple and any other party (Huawei, for instance) because it could then get better results in any future negotiation.

For large corporations enforcing their IP, this is always a major consideration. And it conversely applies to Apple in this case: Apple can show that even if someone does what Qualcomm is doing, they still won't give up their antitrust, contract, patent exhaustion and other claims against Qualcomm.

Potential benefit #3: public relations (PR) and investor relations (IR)

It's been clear over these past two years that no party to patent and patent-related antitrust litigation has ever made a greater PR effort than Qualcomm; and some analysts also find some extreme discrepancies between what Qualcomm tells the courts and what it tells shareholders. PR and IR really drive some of what Qualcomm is doing here, including when they raise hopes of a near-term settlement (apparently Qualcomm's #1 IR objective).

If any of us got slapped with antitrust fines in different places and had to defend ourselves against competition enforcement authorities on three continents in parallel, we'd obviously also like to be able to distract from our own troubles and point fingers at the #1 adversary.

The following quote contained in yesterday's press release contains a false claim, however:

"Two respected courts in two different jurisdictions just in the past two weeks have now confirmed the value of Qualcomm's patents and declared Apple an infringer [emphasis added], ordering a ban on iPhones in the important markets of Germany and China," said Don Rosenberg, executive vice president and general counsel, Qualcomm Incorporated.

The emphasized passage--"declared Apple an infringer"--is the opposite of what the Munich court's own press release says. I'll copy some of the court's original wording below (in German, but I'll translate it immediately):

"Wie der Chip tatsächlich funktioniert, war zwischen den Parteien streitig, nach Auffassung der 7. Zivilkammer aus zivilprozessualen Gründen gleichwohl nicht aufzuklären. Die Kammer hatte ihrer Entscheidung vielmehr aus prozessualen Gründen zugrunde zu legen, dass der Chip so funktioniert, wie die Klägerin es behauptet [...]"

"Die Klägerin konnte in Ermangelung der ursprünglichen Schaltpläne des Chip-Herstellers zu der Funktionsweise des Chips nur anhand eines Teardown-Reports vortragen, den sie anhand eines reverse engineering-Prozesses hatte erstellen lassen. Ein Teardown-Report kann Ungenauigkeiten aufweisen. Die Kammer ging aber aus prozessualen Gründen davon aus, dass der Klägerin mehr nicht abverlangt werden konnte, weil die Schaltpläne des Zulieferers nicht öffentlich zugänglich sind. Die Beklagtenseite bestritt zwar, dass der Chip so funktioniere, wie es die Klägerin behauptete, sah sich aber wegen Geheimhaltungsinteressen ihres Zulieferers außerstande vorzutragen, wie der Chip denn stattdessen funktioniere. Das aber wäre nach Auffassung der Kammer für ein ausreichendes Bestreiten erforderlich gewesen. Weil es an diesem substantiierten Vortrag fehlte, musste die Kammer die Ausführungen der Klägerin zur Funktionsweise des Chips unterstellen und durfte diese nicht näher aufklären."

My translation:

"How the [accused] chip actually works was subject to a dispute between the parties, but [the panel of judges] found for procedural reasons that it did not have to obtain clarification. Under applicable procedural rules, the panel felt forced to base its decision on how [Qualcomm] alleged it worked [...]"

"Qualcomm, for lack of access to [Qorvo's] original chipset schematics, was limited to pleading its allegations based on a teardown report, which it had commissioned on the basis of a reverse-engineering effort. A teardown report can involve inaccuracies. However, the panel came from the assumption that [Qualcomm] could not be required to present more [than the teardown report], given that [Qorvo's] chipset schematics are not publicly accessible. [Apple] denied the accuracy of [Qualcomm's] allegations, but declared itself unable, due to confidentiality obligations, to explain how the chip actually does work. However, such explanation was deemed indispensable by the panel in order for [Apple's] denial to be sufficiently substantiated. In the absence of a substantiated denial, the panel had to take [Qualcomm's] word for it and did not have authority to investigate further."

Note that the court acknowledged it didn't actually clarify how the accused Qorvo chip works (because it concluded the pleadings didn't warrant it), and that Qualcomm's teardown report may have inaccuracies that are inherent to such reports.

So the Munich court stopped far short of declaring Apple an infringer--and even if it had (as it had not), the ITC would not have identified an infringement in a parallel case (with only subtle differences), suggesting that Apple at least had a good-faith basis for what it did.

In the PR/IR context, there's a potential upside for Qualcomm (distraction, display of strength etc.), but also a very significant downside if this goes wrong. In the disputes between Apple and Android device makers (and Microsoft and one Android device maker), no wrongful-enforcement damages were ever determined because the disputes were settled long before. But the Qualcomm situation is so extreme that this could become the first (or at least the first famous) smartphone case in which wrongful enforcement will actually be adjudicated in the end, and that might not end well for Qualcomm's credibility.

Potential benefit #4: Chinese psychology

Qualcomm may hope that the more noise it makes about the German injunction and its enforcement, the more comfortable China may feel that even the leading European patent jurisdiction allows Qualcomm to enforce some patent now. However, there's absolutely no technical overlap with the software patents at issue in China, so this effect should be very limited, if not negligible or zero.

Quick overview of the downside

I can keep this final section short because some of the downside was already discussed in connection with certain limits to the upside, and scenarios in which there might be no upside.

Qualcomm's antitrust worries are still going to be largely the same with respect to Apple, and potentially even worse with respect to the European Commission (which may be very concerned about how Qualcomm seeks to leverage Europe as a battleground, targeting specifically Intel-powered iPhones here), or the Federal Trade Commission (whose trial will begin on January 4), which can see the extreme lengths to which Qualcomm is prepared to go in its patent enforcement against Intel-powered iPhones.

The only way Qualcomm could avoid that everyone sees an anticompetitive intention (and an apparent fear of Apple's antitrust and contract claims) would be for Qualcomm to behave like your average IP enforcer concerned about protecting innovations against free-riding. But Qualcomm behaves precisely in the opposite way, and sophisticated agencies like the FTC and DG COMP will figure it out all too easily:

  • On December 8 I blogged about Qualcomm's proposed findings of fact and conclusions of law in the FTC case. I mentioned that there was a passage that almost raised some interesting estoppel questions. I meant paragraph 420:

    Qualcomm does not seek to prevent others from practicing its cellular SEPs to manufacture or sell components of end-user devices; it neither asserts its SEPs against competing chip makers nor seeks to collect royalties from them.

    The patent-in-suit in Munich is not an SEP, but here, too, Qualcomm simply doesn't seem to be truly interested in efficient IP enforcement: instead of suing the chipset maker (Qorvo), it sues Apple--over Intel-powered devices.

  • As the Munich court explained yesterday, Qualcomm could also just sit on the injunction without enforcing it immediately. In my opinion the most logical thing here for an IP enforcer would have been to do just that--and maybe to begin enforcement after collecting a couple more injunctions against the same devices. In that case Qualcomm could also try, with help from an appeals court, to avoid duplicative (additive) deposits.

  • The functionality at issue here, and the devices at issue (not the 2018 models), are factors that just make a $1.5 billion deposit and all the risks involved (such as having to pay later for the cost of destroying lots of iPhones returned by resellers) totally disproportionate for a minor invention. It's like nuking a kid that's stealing some candy.

Given all the ramifications, and the low likelihood of this bringing Apple back to the negotiating table (it might even make a settlement harder), I think this enforcement effort will ultimately prove to be an inefficient thing for Qualcomm to do at best, and a strategic mistake at worst.

Qualcomm's lawyers are happy about what is technically a win; its PR and IR departments and agencies may also find it useful in the short term; but all things considered, and apart from the fact that the appeals court may soon stay enforcement of the injunction anyway, it looks like it's going to be a Pyrrhic victory.

Share with other professionals via LinkedIn:


Thursday, December 20, 2018

Qualcomm wins envelope tracker patent case against Apple in Munich, but will the injunction ever be enforced?

BREAKING NEWS: Presiding Judge Dr. Matthias Zigann of the Munich I Regional Court ("Landgericht München I") just announced his panel's decisions on, technically, ten Qualcomm v. Apple patent infringement complaints. The key question is now whether a ban of Intel-powered iPhones up to the iPhone X (but not the latest models) will actually be enforced. The court based the infringement holding merely on an allegedly insufficiently-substantiated denial by Apple, not on actual clarification of how the accused chips operate . It's all because Qualcomm alleged.something and Apple couldn't deny it without violating Qorvo's secrets.

For details on what was/is at issue please check out yesterday's preview post. Here's a brief summary of the decisions announced today:

  • The eight complaints in which Qualcomm targeted Apple's Spotlight search will be adjudicated in late January, but given that iOS 12 almost certainly steers clear of infringement, the injunctive relief Qualcomm may or may not obtain them simply won't have any commercial impact--and damages, if ultimately due, would be very limited.

  • In the two cases over a patent on a power-efficient envelope tracker chip design, the court relied on the fact that the court-appointed expert, like the three judges on the panel, couldn't see (given Qorvo's insistence on confidentiality) how the accused chips operate if not the way Qualcomm alleges, and on that basis the court handed down an injunction. This surprises me in light of what Qorvo had told me (and what I found credible and plausible).

    I have no idea how Qorvo's envelope tracker chip achieves the power-saving effect at issue (its inner workings may or may not infringe Qualcomm's patent-in-suit), but it does bear a lot of weight with me that the ITC did not hold a different claim from the same patent family infringed, after evaluating the chipset schematics (which Qualcomm's lawyers prevented from being analyzed in the Munich proceedings) and hearing the testimony of Mike Kay, the designer of the accused chip. I am shocked that Qualcomm's procedural gamesmanship--firstly conducting discovery in the U.S. for the stated purpose of presenting chipset schematics in the Munich court, then making an about-face and asking the German court to rule, and the court-appointed expert to opine, on a basis that's lacking and wanting--has been rewarded. If they have the law and the facts on their side, they deserve to win, but here they wanted--and disconcertingly obtained--a ruling on a basis that I've previously called "evidentiary minimalism" and which would be totally unimaginable in the United States with its far-reaching discovery regime.

    What Qorvo also found unfortunate is that counsel for the formal defendants (with Qorvo being a de facto defendant) didn't have the chance to respond to the court-appointed expert's opinion after further deliberation, though this is (as I know from my own experience here) normally part and parcel of German proceedings involving court-appointed experts. It raises questions regarding the constitutional right to be heard.

    I'm unaware of any commentator who would have expressed greater admiration for the work performed by Quinn Emanuel's Dr. Marcus Grosch. As I predicted, QE Germany has since made it really big. However, I'm just not yet prepared to congratulate them on this shady and fishy win, for two reasons. One, the discrepancy between the outcome at the ITC (based on a robust body of evidence) and the one in Munich makes it quite a possibility that Munich reached the wrong conclusion. Two, because of the way things went wrong here, I wouldn't hold my breath that this injunction will ever get enforced (for more than maybe a negligibly short period of time) as the appeals court--the Munich Higher Regional Court--won't necessarily agree with the court below that a patent holder should be rewarded for not enabling the court and the court-appointed expert to get the benefit of seeing the actual chipset schematics.

    Efficiency is a virtue, and it's often considered a German virtue--but when the interests of justice are disregarded the way it happened here, it's more of a vice. The right balance needs to be struck, and a litigant who starts a discovery proceeding in the U.S., for the stated purpose of obtaining key evidence for a proceeding in Germany, and negotiates a protective order, should not benefit from his counsel making an about-face shortly before trial and declining to accept the terms of that protective order. That's just too much maneuvering. Judge Zigann said today that Qualcomm never promised anything, but it does appear to me that particularly Qorvo reasonably relied on confidentiality. The court says a defendant simply has.to choose: reveal your secrets to deny with specificity an infringement accusation or lose the case.

    I fundamentally--but also respectfully--disagree with the court's procedural approach here, and hope it will be overturned soon. This is way too patentee-friendly, especially when there is a procedural solution through U,S, discovery (best of both worlds).

    Apart from the procedural-evidentiary concern here, there's also the possibility of a U.S. antisuit injunction (first a temporary restraining order) coming down. While the Munich court disagreed with the antitrust-based affirmative defense and ordered an injunction regardless of Qualcomm, the superdominant baseband chipset maker, specifically targeting Intel-powered iPhones, the U.S. district courts with pending Qualcomm antitrust cases (Northern District of California: FTC and consumers v. Qualcomm; Southern District of California: Apple and contract manufacturers v. Qualcomm) may thwart Qualcomm's plan and prevent enforcement of the injunction until the overarching competition issues have been adjudicated in the (earlier-filed) U.S. cases.

    It also remains to be seen whether the European Commission, which launched preliminary investigations into Qualcomm's quest for leverage in Germany, will take this lightly or pressure Qualcomm not to enforce any injunction for now.

Qualcomm brought infringement cases against Apple in the U.S. (primarily betting on the ITC for quick leverage, though that part isn't necessarily working out), Germany, and China. But what started the whole dispute was antitrust enforcement against Qualcomm by regulatory agencies and, through litigation, Apple. We'll still hear news about those infringement cases next year, but what really matters is competition in the baseband chipset market and access to standard-essential patent licenses on FRAND terms. Antitrust enforcement may ultimately even cut through duct tape, and no matter how many infringement cases Qualcomm might win, the competition issues are still the same (if not worse).

Qualcomm's other German lawsuits aren't in good shape. In this post I described the Mannheim situation as a stuttering steamroller, and in October the Munich court threw out another case because it didn't identify an infringement.

But several more German Qualcomm v. Apple cases are pending, with a first hearing in one of them scheduled for late March 2019.

Share with other professionals via LinkedIn:


Wednesday, December 19, 2018

Munich I Regional Court scheduled ten Qualcomm v. Apple decisions for tomorrow (December 20)

This morning, Judge Dr. Anne-Kristin Fricke, in her capacity as spokeswoman for the Munich I Regional Court on civil-law cases, confirmed that the court still plans to hand down decisions in, technically, ten Qualcomm v. Apple patent infringement cases tomorrow (Thursday, December 20) at 2 PM local time (= 8 AM Eastern). The purpose of this preview post is to focus on the forest amid all the trees.

First, in the build-up to the announcement some people got confused and expected "(final) judgments," when "decision" is a broader term that can, in connection with patent infringement lawsuits, mean any of the following (or even other possibilities):

  • a final (though appealable) judgment identifying an infringement of a presumed-valid patent (in that case, an injunction comes down unless that particular remedy is deemed unavailable due to antitrust considerations);

  • a final (though appealable) judgment of non-infringement;

  • an order to stay the case pending parallel invalidation proceedings before the Federal Patent Court of Germany or (in case of very young patents) the European Patent Office;

  • an order to reopen the evidentiary record, which necessitates a new trial date; or

  • an order setting a new decision date.

The public has to stand up for the announcement of a judgment "in the name of the people," but can sit down for (optional) explanations of the reasoning. And if the decision is not a final judgment, everyone can remain seated. There were a couple of Mannheim cases in which I already figured the court was going to order a stay just because we were told to remain seated.

Second, the technically high number of ten cases is due to procedural reasons:

  • German courts always adjudicate only one patent per case (which is good in some ways but does not serve the interests of justice when there are overlapping antitrust issues), even if multiple patents-in-suit are from the same patent family. Five patents will be at issue tomorrow, but four of them are from the same patent family and near-identical. Those four siblings are asserted against an aspect of the Spotlight search. The fifth patent is a chipset (envelope tracker, to be precise) patent.

  • In order to save time, different case numbers are assigned when a plurality of defendants includes at least one EU and at least one non-EU entity. When patent holders bring compülaints against Apple in Germany, they typically sue Apple Inc. (the U.S. parent company), an Ireland-based distribution company, and the German company operating the official German Apple Stores. The Irish and German defendants get one case number (per patent), and the U.S. parent company gets a separate one because service of process takes a bit longer when it doesn't happen within the EU. At least Apple isn't known to hold things up with duct tape...

In practical terms, the pleadings overlap to a huge extent if the same patent is asserted against two sets of defendants (or if multiple patents from the same patent family are asserted), and the courts combine those cases for trial purposes and normally schedule decisions for the same day.

Now that the procedural part has been clarified, we get to what's really important (or not).

In that regard, it's an extreme variation of the 20-80 rule: 80% of the cases on tomorrow's decision list don't really matter. Even Qualcomm's counsel conceded at the September trial that iOS 12 operates differently. He obviously wouldn't concede non-infringement, but what he said was the next best thing. Apparently the difference is simply that iOS 12 displays, exclusively for German customers, a text instead of an icon in one place, and that takes care of the problem. It's preposterous, but, well, it's a software patent.

I expect at least some of the eight Spotlight cases to be thrown out because even older iOS versions didn't seem to practice a claim limitation that some (but not all) of those patents-in-suit have. I don't even believe there is an infringement here of any patent of that family, nor would I predict those patents to survive the parallel invalidation proceedings, but even if the court sided with Qualcomm on some of those eight Spotlight cases, it wouldn't have any non-negligible commercial impact anyway. It wouldn't even be symbolical (in terms of buttressing Qualcomm's claim of Apple free-riding on its innovation), given that those patents (as I explained when I listed them earlier this year) were just acquired, so they don't cover "inventions" made by Qualcomm itself.

I admire the judges and lawyers who can treat such crap seriously with a straight face; I don't know if I could.

All that's really going to be interesting tomorrow is what happens in the pair of cases related to the patent on a power-efficient envelope tracker design (EP2724461). I missed that trial, but thankfully Qorvo, the maker of the accused chip, explained to me what happened and Qorvo's account was plausible and consistent.

In that context, a stay appears unlikely (for all I've been able to research, the court was unconvinced of a serious validity problem), and an outright finding of non-infringement is even harder to imagine since, according to a news agency report, the court-appointed expert witness said he couldn't see how Qorvo's chip wouldn't infringe (though he didn't have access to the chipset schematics because of Qualcomm's tactics).

That means only two possible decisions are likely tomorrow: either the court will identify an infringement on the current basis, or it will conclude that further evidence must be taken. Since I don't know about the inner workings of Qorvo's envelope tracker either, I'll never be able to express an opinion, but what does give me pause is that the ITC (first an Administrative Law Judge, then affirmed by the six-member Commission), after evaluating the chipset schematics (which Qualcomm's lawyers didn't allow to happen in Munich, despite the same law firm--in the U.S.--having negotiated a protective order in a U.S. discovery proceeding for the purpose of use in international litigation) and talking to Qorvo's chipset designer Mike Kay, found a similar claim from the same patent family not to be infringed. That suggests to me that there is at least a possibility of a correct outcome (based on all the facts and testimony) being different from what would result from rewarding Qualcomm's lawyers for evidentiary minimalism that apparently went beyond the dose of opportunism that is par for the course in any litigation.

So, to put it simply, in connection with the envelope tracker patent the only two decisions that are likely would either be an infringement finding on a basis that would fall far short of the one on which the ITC made its decision (and the one that Qualcomm's German counsel had originally agreed to but later walked back on) or a procedural order to reopen proceedings and take the necessary time and evidence to arrive at a well-considered decision. In the latter case, Qualcomm couldn't complain about the delay because its German counsel could--instead of the about-face they made--simply have signed the U.S. protective order and presented the chipset schematics to the German court and the court-appointed expert witness.

If that Qorvo chip was deemed to infringe without all the facts having been on the table, I would feel particularly sorry for the Qorvo guys considering that Qualcomm isn't actually interested in that company (with which it even partners in some ways). Qualcomm is seeking leverage over Apple in order to get rid of the antitrust and contract law challenges pending in the Southern District of California and foreign jurisdictions, and in order to force Apple to drop Intel as a baseband chipset supplier. The latter would potentially enable the Munich court to deny injunctive relief. It's hard to avoid a patent injunction in Germany if liability has been established, but in this case the technical nexus between the relevant patent assertion and Intel's baseband chip is infinitely closer than in, say, the Spotlight search cases.

Also, I don't know how Judge Koh in the Northern District of California and Judge Curiel in the Southern District of California would view such an injunction just before they're going to adjudicate some fundamental issues regarding Qualcomm's allegedly anticompetitive conduct. Maybe we'd see an antisuit injunction aiming to prevent Qualcomm from enforcement in Germany.

In other news, Ted Boutrous, a lawyer representing Apple's contract manufacturers against Qualcomm in the San Diego federal lawsuit (and whose firm, Gibson Dunn, has represented Apple in other cases, such as against Samsung), gave Bloomberg an interview on the Chinese situation. In China, like in the German Spotlight cases, Qualcomm is apparently asserting acquired (not homegrown) patents, and Mr. Boutrous says there's no way Qualcomm could realistically get the leverage there that it seeks. For lack of access to first-hand information on the Chinese cases, I can't form an opinion at this stage, but the statements made in that interview come across as pretty plausible given how software patents are usually worked around by alleged infringers.

Share with other professionals via LinkedIn:


Tuesday, December 18, 2018

Funny photo: former Qualcomm president duct-tapes gate to avoid testifying at January FTC trial

The following picture shows the pedestrian gate of former Qualcomm president Derek Aberle's San Diego home. Even high-tech execs resort to no-tech means--good ol' duct tape, in this case--when trying to avoid having to testify in court (this post continues below the image):

Last week the Federal Trade Commission requested the court's permission to serve trial subpoenas on former Qualcomm president Derek Aberle and former Qualcomm vice chairman Steven Altman by mail after the two "repeatedly evaded attempts to personally serve subpoenas seeking their apperance at [the January 2019 FTC v. Qualcomm] trial."

In order to document those fruitless efforts, the FTC attached, among other things, a report by a San Diego process server, Mr. Manfred Arnold Briese, Jr.:

"12/10/2018 at 4:45- Arrived at location to relieve previous server [so they were basically camping outside the property]

5:40pm-Black Land Rover SUV arrived with female adult and minor male

5:45pm- Minor male came out to pedestrian gate and duct taped it closed [document by the photo shown further above]

6:26pm- Black SUV left residence and returned at 6:35pm.

10:00pm-No activity. Left location."

[and a similar report for the following day]

The FTC also tried to serve the subpoenas through Cravath, a law firm representing Qualcomm but also Messrs. Altman and Aberle at earlier depositions. However, Cravath denied being authorized to accept documents on behalf of Messrs. Altman and Aberle, pointing to the fact that Mr. Aberle is part of a group formed by former Qualcomm chairman Paul Jacobs, trying to raise funds to take over (and privatize) Qualcomm.

Mr. Aberle apparently concluded that testifying at trial wouldn't be in his interest. He doesn't want to do anyone a favor by showing up. He wants to be left alone. So he's hiding behind a duct-taped gate--or he's somewhere else, but the people currently living on his property are hiding.

What won't help him avoid or delay his testimony is a potential federal government shutdown. Judge Koh handed down an order yesterday according to which the trial will go forward in any event.

Share with other professionals via LinkedIn:


Friday, December 14, 2018

Judge Koh rejects Qualcomm's bid to introduce new evidence related to Intel's success with Apple, Qualcomm's new 5G agreements

We're now only three weeks away from the Federal Trade Commission v. Qualcomm antitrust trial in the Northern District of California, and Judge Lucy H. Koh has again asserted her authority and demonstrated her unrelenting focus on the facts that really matter. On Thursday, she flatly rejected an 11th-hour attempt by Qualcomm to delay or derail the FTC's push for injunctive relief (the FTC is seeking to "redress and prevent recurrence of Qualcomm's conduct") by introducing evidence more recent than March 30, 2018, and at the end of her sharp ruling told Qualcomm one more time that "[t]he January 2019 trial will address both liability and remedy" (this post continues below the document):

18-12-13 Order Denying Qualcomm's Request Re. Post-discovery Events by Florian Mueller on Scribd

In late October, Qualcomm brought a motion to introduce evidence related to post-discovery (i.e., post-March 30) events that the wireless chipset maker believes weigh against the FTC's pursuit of an injunction:

  • Intel's growing business with Apple (a design win related to the entire range of next-generation iPhones, though Intel, as the ruling notes, already had a chipset supply deal in place with Apple well before the discovery cutoff date) and

  • some 5G patent license agreements Qualcomm has concluded more recently, arguing that since it doesn't make any 5G chips (yet), it can't leverage a chipset monopoly ("no license-no chips" policy) in connection with 5G. This part did not refer to Qualcomm's "new deal" with Samsung (which it may have to renegotiate anyway), which was concluded earlier this year, well ahead of the discovery cutoff date.

In formal legal terms (FRE 403), Judge Koh concluded that allowing Qualcomm to take and present further evidence would unfairly prejudice the FTC, and that such prejudice would be outweighed by the alleged probative value of anything Qualcomm would inject into the case. In her efforts to balance different considerations, the judge wasn't persuaded by precedents Qualcomm pointed to, especially not by other cases in which several years had passed between a discovery cutoff date and a trial. She also points to the fact that she actually did allow various out-of-time depositions because of witnesses's travel and for similar reasons. And she notes that there will always be something that happens post-discovery, so no case can ever go to trial (at least not to one that would be fair to the opposing party) unless courts exercise their discretion at some point to close the door. And in this case, the deluge of motions that Judge Koh faced (her ruling doesn't say "deluge," but contains an impressive list) between the cutoff date and the trial doesn't suggest that the period between the cutoff date and the trial was longer than what the court really needed. While I've sometimes had different opinions, especially on economic issues and the (in)validity of patents-in-suit, I can say from my vantage point as a litigation watcher that I'm not aware of any U.S. judge who would be more efficient than her.

The key factor here is that the FTC is entitled to injunctive relief if there is a risk of Qualcomm's (mis)conduct recurring--regardless of whether, as Qualcomm argues, anything may have recently changed about its market power. Judge Koh points to case law that makes this principle crystal clear. If Qualcomm had wanted a different outcome, it would have had to show that its conduct had changed, and even a change of conduct wouldn't necessarily have meant that there was no more risk of recurrence.

As Judge Koh recalls, Qualcomm's desire to avoid a decision on injunctive relief after the January 2019 trial is old news. Qualcomm originally sought to bifurcate the trial (one trial on liability followed by a separate one on remedies).

The fact that Intel's deal with Apple is not new per se (though the scope of that partnership may have expanded lately) diminishes the probative value of anything that Intel and Apple may have done together, or may have decided to do together, since late March. As for the 5G part, Judge Koh doesn't mention Qualcomm's deal with Samsung, but the fact that this modification of an agreement with the world's largest smartphone maker (at least by some measure) occurred prior to the cutoff date also suggests that whatever other deals were struck in the second or third quarter of 2018 won't present a fundamentally new situation. And it could be that Judge Koh already sees a certain likelihood of Qualcomm being ordered to renegotiate patent license agreements at the end of this. In that case, the probative value of such contracts might be zero.

This latest order is not the first major pretrial ruling that doesn't bode well for Qualcomm's defensive efforts:

The exclusion of evidence relating to post-cutoff date events is the latest in a string of significant decisions made and positions taken before the upcoming trial.

Assuming that Qualcomm loses the trial (and I can't imagine it would get away unscathed), it will then have to set priorities for an appeal, and I guess--though this is now a second-degree hypothetical as it already presupposes a certain outcome of the trial--Qualcomm is fairly likely to tell the appeals court that it lost only because some relevant post-discovery events weren't considered and to request a reversal or at least a retrial. But Judge Koh obviously considered this scenario and structured her ruling accordingly.

There most probably won't be any spectacular developments in the U.S. Qualcomm cases between now and the end of the year. Maybe we'll hear something more about the software patent enforcmeent dispute in China (to be fair to Qualcomm, when Apple enforces its own patents, it doesn't easily accept that a workaround is indeed a workaround). And I'm extremely interested in what the Munich I Regional Court will decided next Thursday (December 20). "Cutoff" for evidence also plays a role there, but it's different in three major respects: a different jurisdiction, where the constitutional right to be heard quite often weighs in favor of reopening the record; it's a patent infringement, not antitrust, case; and what Qorvo credibly explained to me suggests that Qualcomm originally led the court and Apple to believe that the schematics of the accused chipset would be discussed with the court-appointed expert at trial, but then made an about-face and opted for evidentiary minimalism, which is why in that case I believe the plaintiff would not be unfairly prejudiced by taking a bit more time to put facts on the table that, based on the ITC decision (first an Administrative Law Judge cleared Qorvo of infringement, then the six-member Commission threw out Qualcomm's petition to review that finding), have the potential to be outcome-determinative in the defendants' favor.

Share with other professionals via LinkedIn:


Thursday, December 13, 2018

ITC decides to review initial determination that Intel-powered iPhones infringe one Qualcomm patent

Yesterday afternoon the United States International Trade Commission ("USITC" or just "ITC"), a U.S. trade agency with quasi-judicial powers, issued a determination to review meanwhile-retired Administrative Law Judge (ALJ) Thomas B. Pender's initial determination according to which Intel-powered iPhones infringe one Qualcomm patent (U.S. Patent No. 9,535,490 on "power[-]saving techniques in computing devices").

A headline I read this morning suggested that this is a "second shot" for Qualcomm to prevail, which may be attributable to spin doctoring and is fake news just like a story according to which Apple CEO Tim Cook was preparing a 2020 presidential run (paid for by Qualcomm, presumably in an effort to adversely impact Mr. Cook's relationship with President Trump). The determination to conduct a review is actually a setback--not progress--for Qualcomm if, which gained nothing whatsoever yesterday if one analyzes carefully what the ITC actually decided and compares it to what it could have decided in the alternative.

First, as I explained on November 2, a fundamental distinction must be made between an initial determation (ID) on the merits (whether a valid patent is infringed) and a recommended determination (RD) on remedy and bond. The ALJ issues both simultaneously, but the ID (or any given part of it) becomes the final Commission determination if the Commission--the six-member decision-making body at the top of the organization--decides not to review it, while the RD is just what its name says (a recommendation) and the Commission must make a determination on those questions whether or not anyone files a petition (in fact, it's not even possible to formally "petition" for a "review" of the RD because only decisions, not recommendations, can be "reviewed"; a recommendation can, at best, be adopted).

The public-interest inquiry is only relevant to the remedy part:

  • Yes, ALJ Pender recommended not to order an import ban over the patent he deemed infringed, as he viewed Qualcomm's related ITC complaint as part of an anti-Intel campaign with serious anticompetitive implications (for a closer look at his findings, see this post).

  • Yes, the Commission yesterday requested answers to five public-interest questions (vs. nine related to the merits), and it's procedurally possible that the Commission ultimately reaches a different conclusion than ALJ Pender.

  • But no, this does not mean that the Commission has taken a procedural decision in Qualcomm's favor (such as by granting a petition to review) or identified any potential reasons to disagree with ALJ Pender. As I've explained repeatedly, and as yesterday's document actually also states (though not as clearly as this post), the Commission must evaluate the statutory public-interest factors. It couldn't just say: "we affirm the ALJ". Even if the Commission wholeheartedly agreed with the recommendation, it would have to make its own determination, and would have to request input from stakeholders.

While the parties couldn't even petition for a "review" of the RD on remedy and bond, the Commission requested public interest statements in the build-up to yesterday's determination. The implications of an anti-Intel ban on U.S. competitiveness in 5G were put front and center by Apple's submission, and yesterday's public-interest question D also refers to "the alleged exit of Apple's chipset supplier from the market for 5G technology."

What remains to be seen now is whether the public-interest questions will even be reached by the Commission. Unless Qualcomm prevails on at least one patent claim, there obviously won't be a remedy.

With respect to the merits, the Commission

  • granted Apple's petition to review the liability finding with respect to the '490 patent, with Apple now getting the proverbial second bite at the apple to defeat the complaint (in fact, a second round of three bites: the Commission's nine questions involve infringement, liability, and domestic industry, any one of which would be a "get out of jail free" card for Apple on its own), but

  • denied Qualcomm's (conditional) petition to review the holdings that Apple did not infringe the other two patents ALJ Pender ruled on.

One key difference here is that Apple's petition for review was unconditional (Apple definitively asked the Commission to overrule ALJ Pender on liability), while Qualcomm's petition was conditional. Qualcomm wanted the Commission to adopt the ALJ's initial determination on the merits, but also wanted second bites at the apple (in this case, at "the" Apple with a capital A) if a review was going to take place. What has happened now is that the ITC decided to review the part that is good for Qualcomm (and can only get worse for Qualcomm, not better), but declined to review the parts where Qualcomm could have made further headway.

This also means that the Commission won't review the ALJ's finding that a certain Qorvo chip in Intel-powered iPhones doesn't infringe a Qualcomm patent that has a European "sibling" that is at issue in a case the Munich court will decide on next week.

Obviously, Qualcomm's conditional petition for review looked weak from the start. A party that believes there's a good chance of prevailing on one or two additional patents simply requests a review on an unconditional basis. Many other ITC complainants have sought to widen, broaden and deepen their ITC wins, and in all the cases I saw, except for this Qualcomm case, those petitions were unconditional. Theoretically, a party's desire to save time and rather get one bird quickly than two or three birds later could result in a conditional petition for review even if the case was strong. However, we're just talking about a couple of months here that Qualcomm could have saved in the best case--not much given that this dispute started almost two years ago. The Commission determination has been scheduled for February 19, 2019.

So by now, there's only one patent in the case, and Apple has four different chances to avoid an import ban, and needs to prevail on just one while Qualcomm must overcome all four hurdles:

  1. infringement (a matter of claim construction, and the Commission is interested in claim construction arguments),

  2. validity (the Commission appears particularly interested in one prior art reference, but also in a second one that might be relevant to an obviousness analysis),

  3. domestic industry (here, claim construction also comes into play), and

  4. the public interest in thwarting a market monopolization campaign.

Meanwhile, media reports suggest there's a bitter enforcement fight raging in China, where Qualcomm just secured two preliminary injunctions against Apple (from what I hear, there had only been two other patent PIs in Chinese history before these two), but Apple argues that iOS 12, which the court didn't look at, doesn't infringe. The patents in that case are software patents, and iOS 12 contains at least one other workaround related to a Qualcomm patent-in-suit. As Apple's and Microsoft's patent enforcement efforts against Android devices showed years ago, it's very hard to find a software patent that is broad enough so it can't just be worked around easily (and in ways that end users often don't even notice), yet not so broad that it would be held invalid. Anyway, I only have indirect information (media reports) on the Chinese situation and will continue to focus on jurisdictions in which I can conduct primary research (analyzing documents and attending hearings/trials/announcements).

Share with other professionals via LinkedIn:


Saturday, December 8, 2018

Qualcomm fears being required to renegotiate patent license agreements with Samsung, many others

This is the first post, and probably not the last, in which I'll discuss some interesting information I found in the Federal Trade Commission's and Qualcomm's proposed findings of facts and conclusions of law with a view to next month's San Jose trial. Qualcomm's filing is more than twice as long (157 pages) as the FTC's submission (71 pages), but Judge Lucy H. Koh will decide strictly based on the law and the facts, so this antitrust case is not going to turn into a battle of matériel. The litigation departments of government agencies are outnumbered by private-sector litigants' armies of lawyers all the time, but quite often they prevail nevertheless.

Here are the two filings (this post continues below the documents):

18-12-06 FTC's Proposed Findings and Conclusions by Florian Mueller on Scribd

18-12-06 Qualcomm's Pro... by on Scribd

Those proposed findings and conclusions serve as a roadmap for the upcoming trial. There may still be some surprises, but by and large the points the parties seek to drive home are clear now, except for those heavy redactions, which I wish someone could fight the way Reuters, MLex, the First Amend Coalition and others attacked excessive sealings in Apple v. Samsung years ago (and I had to do the same at the Federal Patent Court of Germany; by the way, I'm presently involved in a new access-to-documents dispute there, with a notorious privateer).

Section III.I of Qualcomm's filing is one of the most outrageous examples of overredaction. That section is meant to justify Qualcomm's "no license-no chips" policy, which is central to the case. Under the headline "Qualcomm's Practice of Not Selling Chips to Unlicensed Companies Has Valid Business Justifications," paragraph 245 says:

"Qualcomm's practice of not selling chips to unlicensed OEMs is based on the following considerations:"

And then every single one of those business considerations is blackened out. 100%. It's hard to imagine that those redactions are reasonable, and maybe the court won't approve them. Or we'll find out at trial.

Fortunately, the paragraph that is most interesting in commercial terms is public:

"765. Moreover, the FTC seeks an order requiring Qualcomm to 'renegotiate . . . license terms' with all OEM licensees. (Joint Pretrial Statement at 2; FTC Interrogatory Response at 6.) Not only would such an injunction require renegotiation in markets where Qualcomm is not even alleged to have market power, such as WCDMA or non-'premium' LTE, but it is also overbroad insofar as it requires renegotiation of agreements no matter what the terms and no matter whether any anticompetitive harms caused or resulted from those agreements. Such an injunction would be tailored neither to the markets nor harms at issue, including because it would apply to agreements entered into outside the 2011-2016 timeframe for which the FTC has presented evidence."

The passage on the requested relief that Qualcomm's filing refers to says this:

"Require Qualcomm to negotiate or renegotiate, as applicable, license terms with customers in good faith under conditions free from the threat of lack of access to or discriminatory provision of modem chip supply or associated technical, software, or other support;"

It's obvious why Qualcomm doesn't like this, but it's inevitable: if someone violates antitrust law and as a result of such behavior (including, but not limited to, the "no license-no chips" policy) imposes supra-FRAND royalties on others, renegotiation is the only way to redress the balance and fix the problem.

As the final sentence of paragraph 765 of Qualcomm's proposed findings and conclusions shows, they complain, among other things, about the requested injunction not being limited to the 2011-2016 period. The most significant agreement that is still in force and effect and that Qualcomm concluded outside that timeframe is presumably its early 2018 new deal with Samsung. Even if the FTC didn't present evidence that is specific to that timeframe, there can be no doubt that the 2018 Qualcomm-Samsung agreement came into being under the same problematic circumstances--simply because things will only get better if and when the FTC prevails (or if and when private parties prevail, with the Apple v. Qualcomm trial being scheduled for mid-April).

Qualcomm's proposed findings and conclusions do indicate some changes to their business policies after that timeframe, and they seek to leverage those changes so as to argue that the prospective remedy of injunctive relief wouldn't be warranted only on the basis of past behavior. The risk of recurrence is obviously key. But what Qualcomm changed (such as capping its 5% patent royalty demand at a device price of $400) is either unrelated to, or at least falls far short of, what the FTC is tackling here and seeking to redress and to prevent from happening again.

Actually, to the extent Qualcomm's refusal to extend SEP licenses on FRAND terms to rival chipset makers like Intel was a breach of contract (as Judge Koh has determined on summary judgment), that fact alone means that the economics of certain deals, such as the one with Samsung, would have been different if Qualcomm had behaved the way the court thinks it should have, and if a licensee like Samsung could also have decided to simply purchase baseband chips from Intel, or to license Qualcomm's SEPs at the chipset level (Samsung's Exynos division) instead of at the device level.

The requirement to renegotiate license terms would affect Qualcomm in two ways. Besides the direct implications of having to modify existing agreements in favor of implementers, a judge or jury tasked with determining whether or not Qualcomm complied with its FRAND licensing obligations will also look at (among other things) Qualcomm's other license agreements. It's a safe assumption that Qualcomm's recent deals, such as the one with Samsung, were optimized for the purposes of the forthcoming Apple trial. However, should those agreements be null and void in the sense that they must be renegotiated as per a court order, then they don't serve as useful points of reference for any FRAND-compliance analysis in any jurisdiction.

Again, there's a lot more in the 220+ pages Qualcomm (two thirds) and the FTC (one third) filed on Thursday, such as an interesting passage that could backfire against Qualcomm in connection with promissory estoppel (and possibly have novel implications for exhaustion), but it's simply too much for a single blog post.

Share with other professionals via LinkedIn:


Friday, December 7, 2018

When defending against standard-essential patents, beware of standard-essential utility models

While "standard-essential patents" (SEPs) is one of the most common terms in the tech sector, it would sometimes be more accurate and inclusive to refer to "standard-essential intellectual property rights" (SEIPRs). That collective term would include both SEPs and SEUMs: standard-essential utility models. Utility models are a German specialty, basically a second-class type of patent with a shorter term but instant registration (no substantive examination). This blog covered a utility model case years ago when Apple asserted a slide-to-unlock utility model against Samsung in Germany; that case got stayed over validity concerns and never went anywhere. Beyond German utility models, the collective term "SEIPRs" would cover any other IPRs that may exist in other jurisdictions and are like patents, but aren't called patents.

Yesterday I went to the Munich I Regional Court to watch a standard-essential utility model case, Netlist v. SK Hynix and HP, over German utility model no. DE2020100185017, which was derived last year, in preparation of this lawsuit as counsel for Netlist explained, from a pending European patent application, EP2454735 on a "system and method utilizing distributed byte-wise buffers on a memory module." From what I've been able to find out, this patent was declared essential to a JEDEC memory standard.

An EPO patent examiner rejected the application, though Netlist is still trying to persuade the EPO to grant a patent. But in parallel to that effort, they quickly took out a utility model, with claim language drafted specifically for the purposes of the lawsuit against SK Hynix and HP, and sued in Munich.

The previous Munich trial I reported on took 12 hours; yesterday's trial took only about four. No would-be witness was sitting on the hard wooden seats outside.

Netlist's counsel in this action, Fieldfisher's Benjamin Grzimek, was mentioned on this blog about five years ago when Unwired Planet, an Ericsson privateer, sued a bunch of mobile device makers. A filing made by SK Hynix with the ITC suggests Netlist is also a patent assertion entity (unsurprising after I had noticed the extent to which they emphasize their patent infringement cases on their corporate website):

"Netlist does not design or manufacture JEDEC compliant DDR4 RDIMM or LRDIMM, and it never has. It does not compete with Respondents. And, Netlist freely admits that it made no technical contribution to any JEDEC standard. Others contributed the technologies JEDEC adopted into its standards, which JEDEC compliant products implement. Indeed, in years past when Netlist designed and sold memory modules, its business model was to sell non-JEDEC compliant memory modules; these are the products, according to Netlist, from which the patents arose. Tellingly, despite having litigated the patents and defended IPRs, Netlist did not even disclose four of the five patents asserted at the [ITC] hearing to JEDEC as potentially essential to the standard until years after the patents issued, and just months before filing its [ITC] complaint. All of this reflects a shift in Netlist's overall strategy. After operating at a loss for all but one year of its existence, Netlist now touts that '[t]he Company is focused on monetizing its patent portfolio' and that it 'plans to pursue an intellectual property-based licensing business in which it would generate revenue by selling or licensing its technology, and it intends to vigorously enforce its patent rights against infringers of such rights.'"

"Consistent with this strategy, Netlist has secured third-party funding for its litigation campaign against SK hynix, which is by no means limited to the two investigations in the ITC. Netlist recently announced that it 'obtained outside investment to finance the legal fees and costs of its legal action against SK hynix” from TR Global Funding V, LLC."

The defendants in the Munich utility model are represented by IP and antitrust litigators from Bardehle Pagenberg led by Professor Peter Chrocziel (who has been named German IP litigator of the year twice and was mentioned numerous times on this blog because of his past work for Microsoft and Apple), with rising star Dr. Anna Giedke arguing non-infringement for the most part. Interestingly, even though Bardehle has many patent attorneys itself, including some of the most well-known ones in Europe, two patent attorneys from Samson & Partner (another top-notch firm I mentioned often because of its work on behalf of Nokia and, such as in the pending Qualcomm cases, Apple) represented SK Hynix and HP yesterday: Dr. Georg Jacoby and Dr. Robert Baier. The involvement of Samson's patent attorneys doesn't imply anything negative for Bardehle's great patent attorneys; if anything, it speaks to the strength of Bardehle's litigation team.

The court will announce a decision at the end of January, and Presding Judge Dr. Matthias Zigann said at the end of the trial that he can't indicate an inclination before a post-trial conference with his two side judges. If I had to make a guess, I'd expect the case to be dismissed or stayed. Defendants made a pretty good "squeeze" case, where the asserted claim would either have to be construed too narrowly to support an infringement finding or, in the alternative, the claim would be too broad to be valid. A patent attorney from the Bosch Jehle firm, representing Netlist, showed a computer animation meant to distinguish the claimed invention from the prior art, but patent attorney Dr. Baier dismissed its core--the idea of taking CAS latency into account-as a "triviality" and, while Judge Dr. Zigann didn't say what he thought of it, I wouldn't bet on Netlist's chances here.

Judge Dr. Zigann noted at the start of the trial that the utility model and its underlying patent specification lack clarity. He explained that "the upside for you is that counsel can read anything into it; the downside is that so can the Court."

There are two important lessons to be learned:

  1. Reference was made to a U.S. case in which one or more defendants sought or even obtained (at some point, however, that case got stayed by stipulation) a U.S. antisuit injunction barring Netlist from the enforcement of injunctive relief over SEPs--but if Netlist prevailed in the case tried by the Munich court yesterday, there would be an argument over whether an SEUM falls within the scope of that injunction. I believe SK Hynix and/or HP would have a strong case, given that this SEUM was simply derived from a pending SEP application. But there could be an enforcement dispute, and there would be a risk of a court saying that if they wanted to preclude Netlist from enforcing utility models, they should have requested a differently-worded antisuit injunction. And that could give Netlist leverage, even if perhaps only for a short period.

  2. While utility models don't enjoy any presumption of validity (for lack of substantive examination), there can be situations in which a (final or non-final) rejection of the underlying patent application by the EPO actually suggests to a German court that the utility model is likely valid. In yesterday's case, that doesn't appear to be an issue. But there have been utility model cases in Germany in which an EPO examiner based a rejection on a theory that doesn't apply to German utility models. The differences between the two jurisdictions are limited and sometimes subtle, but there are some and they can prove outcome-determinative, resulting in a utility-model injunction while a patent couldn't be enforced.

Implementers of standards should always bear the additional risks arising from utility models in mind, especially when there is a threat of litigation, or already an ongoing one. You need an invalidation strategy that will not only work against a European patent or patent application, but also take down a utility model. And you shouldn't forget about utility models when crafting antisuit injunction motions.

Share with other professionals via LinkedIn:


Monday, December 3, 2018

FTC seeking to "redress and prevent recurrence of Qualcomm's conduct" through antitrust injunction

The FTC and Qualcomm once intended to settle the antitrust litigation pending before Judge Lucy H. Koh in the Northern District of California by November 14, but we're now just one month and one day away from the trial date and no agreement has been reached. But there's been tremendous progress in the form of Judge Koh's recent summary judgment order on Qualcomm's obligation to extend standard-essential patent (SEP) licenses to rival chipset makers such as Intel.

On Thursday, the FTC and Qualcomm filed a joint final pretrial statement, which outlines (on page 4 of the PDF document) the remedies the FTC is going to fight for (this post continues below the document):

18-11-29 FTC v. Qualcomm Jo... by on Scribd

Let's look at the injunctive relief sought by the FTC item by item:

  • "Prohibit Qualcomm from conditioning the supply of modem chips on a customer’s patent-license status"

    This would be the end of Qualcomm's "no license, no chips" policy. Qualcomm's going to argue that it shouldn't have to tolerate that its chipset customers infringe on its patents, but as a monopolist it's not allowed to engage in tying and, regardless of market share, patents embodied by its chip are simply exhausted. Last year's Lexmark ruling by the Supreme Court makes this pretty clear, one would think.

  • "Require Qualcomm to negotiate or renegotiate, as applicable, license terms with customers in good faith under conditions free from the threat of lack of access to or discriminatory provision of modem chip supply or associated technical, software, or other support;"

    This, too, would prevent Qualcomm from leveraging its chipset market position when negotiating license terms.

  • "Require Qualcomm to submit, as necessary, to arbitral or judicial dispute resolution to determine reasonable royalties and other license terms should a customer choose to pursue such resolution;"

    I've warned against arbitration on unfair terms on various occasions and in different ocntexts (most recently, Huawei v. Samsung). What's good here is the requirement that a Qualcomm customer would "choose to pursue such [dispute] resolution." This way, customers preferring to have an Article III court make a FRAND determination can't be forced to arbitrate.

  • "Require Qualcomm to make exhaustive SEP licenses available to modem-chip suppliers on fair, reasonable, and non-discriminatory terms and to submit, as necessary, to arbitral or judicial dispute resolution to determine such terms;"

    The second reference to arbitration must be seen in light of the first one: it's up to the licensee whether the matter is arbitrated or litigated.

    This prayer for injunctive relief goes beyond the summary judgment victory the FTC has already scored. As the FTC clarified in August, the summary judgment motion was merely about contract interpretation, not about a general requirement on antitrust grounds. Now they're shooting for a bright-line rule.

  • "Prohibit Qualcomm from discriminating or retaliating in any way against any modem-chip customer or modem-chip supplier because of a dispute with Qualcomm over license terms or because of a customer's license status;"

    The intention behind this one is good and clear. However, Qualcomm's criticism of the FTC's requested relief lacking specificity is not entirely unfounded with respect to this broad and vague terminology: "retaliating in any way" could also involve legitimate forms of intellectual property enforcement. Maybe the FTC will provide greater clarity.

  • "Prohibit Qualcomm from making payments or providing other value contingent on a customer's agreement to license terms;"

    This one is meant to close a loophole: Qualcomm has previously entered into rebate/kickback agreements in order to get companies to accept contract terms with anticompetitive implications.

  • "Prohibit Qualcomm from entering express or de facto exclusive-dealing agreements for the supply of modem chips;"

    Here, the FTC is seeking to prevent Qualcomm from monopolization through contracts.

  • "Prohibit Qualcomm from interfering with the ability of any customer to communicate with a government agency about a potential law enforcement or regulatory matter;"

    This issue is well-known because of Apple's dispute with Qualcomm. Qualcomm promised Apple some rebates, but then withheld a billion-dollar amount just because of Apple allegedly having violated a "business cooperation and patent agreement" (BCPA) by talking to regulators, which Apple says was in some cases merely about responding to questions from competition enforcers and in the remaining cases happened after the BCPA expired. Apple argues that it didn't violate the agreement while it was in force, and didn't emphasize the policy implications of such contract terms too much; fortunately, the FTC is now trying to establish a general rule that Qualcomm's customers must remain free to talk to regulators,even proactively as complainants.

  • "Require Qualcomm to adhere to compliance and monitoring procedures and appropriate 'fencing in' provisions, including but not limited to a potential firewall between patent licensing and chip personnel;"

  • I'm not sure this would work in practice. In either division, employees would likely try to optimize the result for Qualcomm as a whole.

  • "Impose any other relief that the Court finds necessary and appropriate to redress and prevent recurrence of Qualcomm's conduct"

    This is just meant to give Judge Koh greater flexibility. What it means, and whether it will give rise to any particular injunction after the bench trial, remains to be seen.

These prayers for injunctive relief suggest that the trial is going to be extremely interesting, and what will happen afterwards may be very significant. Parties sometimes settle on the eve of the trial, so it's not certain yet that the trial will happpen. But, at a minimum, any settlement (consent decree) would have to be viewed against the background of the above proposals. If the FTC settled on a basis that would fall far short of the above, the agency's credibility would be compromised...

In other news from last week, Judge Gonzalo P. Curiel scheduled the Apple v. Qualcomm trial in the Southern District of California to begin on April 15. It's a safe assumption that Judge Koh, who is known to work both smart and hard, will hand down her decision on the FTC case before the San Diego trial between Apple and Qualcomm. If Qualcomm's "no license, no chips" policy and other tying and various forms of threatening or discriminatory behavior came to an end and/or if Qualcomm's contract terms prohibiting antitrust complaints were held illegal ahead of the trial in Southern California, Judge Curiel could presumably streamline his case in different ways, narrowing the issues to be put before the jury.

Share with other professionals via LinkedIn: