Landmark Settlement of Apple v. Qualcomm Gives U.S. Cellular Industry a Way Forward
A legal showdown that had reached maximum intensity ended peacefully last week. The suit and countersuit between Apple — the giant smartphone maker — and Qualcomm — the supplier of much of the technology that makes a smartphone smart — settled a bit beyond the proverbial and literal courthouse steps this past Tuesday.
A bit beyond because, even though the key parties — not the jurors, the press, nor the pubic, but the attorneys, management on both sides, and even Judge Gonzalo P. Curiel of the U.S. District Court of Southern California, in whose forum the case was going forward — knew the settlement was afoot, the judge insisted that unless all the ‘t’s were crossed, opening arguments should begin.
Judge Curiel is now a rockstar. Understated as ever, this obscure jurist on the California bench was elevated to international fame when then-candidate Donald Trump took reputational aim at him in an effort to deflect culpability in the Trump University case, which later settled with a fine. In the weird logic of branding in the United States, that ethnic smear worked in favor of Judge Curiel. Go figure.
It was a pleasure to sit in his court and watch his affable but serious demeanor as he gave the jury their initial instructions. The courtroom was tiny, and the gallery limited. I had lined up more than an hour before the courthouse opened with the most dedicated press and managed to get one of the coveted seats inside (later arrivals were shunted off to an overflow room in the next building over, where they watched the proceedings via video). When there was a small panic over whether actual seats would be available for everyone who had waited, one of the journalists, referring to the 10 dedicated chairs for press, snapped at me, “These are for real journalists!”
She was perfectly right to question my credentials. I am an industry analyst and a Qualcomm partisan. I’ve had a paid consulting relationship with Qualcomm for a decade. And my relationship with Apple ended abruptly fifteen years ago, when I felt the metaphorical back of Steve Jobs’s hand over my failure to adhere to message. So, my sentiment could be presumed to be pro-Qualcomm a priori. However, this has been one of my more joyous engagements, since my personal views on these particular matters are so closely aligned with Qualcomm’s. So I report this, not as a journalist, but as an observer with skin in the game.
Turns out there were a number of “seat warmers,” who were holding places for more senior functionaries, who could then show up just in time or not at all. Thus, a seat opened up across the way from the real journalists, and I got to see the action live. It was pretty exciting.
After the eight jurors and an alternate were seated, the parties went through the charade of opening statements, even though they knew it would all be over in a few hours. But these initial arguments were significant because, even though, as Judge Curiel pointed out to the jury, lawyers’ statements are opinions rather than fact, Qualcomm’s attorneys were able to place actual evidence into the record by referring to prepared slides containing factual information. As the intrepid reader will see, it was particularly crucial to get some of this on the record because once the case settled, all these juicy tidbits would be removed from view, and we, the public, would never learn the true facts of the case, only the meager crumbs in the anodyne agreement.
In the cheaper form of journalism, stories are often written as a tale of winners and losers, and I did see a bunch of those narratives out there in the immediate aftermath: Qualcomm wins big; Apple loses big. But the more discerning reader will realize that, although Qualcomm did win big, Apple did not lose. Yes, Apple has to write a check to Qualcomm for something resembling the $8 billion in patent royalties the iPhone maker has withheld over the past two years, and that certainly comes as a relief to Qualcomm, which had throttled expenses trying to cope with this rather large hole in the boat. But Apple has more cash than most sovereign countries, and a lot less debt. So, it was really just a small nick. And Apple gets to use Qualcomm’s considerable technology portfolio. And within that portfolio, the most important elements to Apple relate to next-generation cellular communications.
Yes, in the end it came down to the 5G modem. Apple had thrown its communications lot in with Intel, which had wiffed round one, failing to produce a viable radio, and was in danger of wiffing round two. Round three was set to be Apple’s own modem, but, as the timeline drew out, Apple appeared to be in greater and greater danger of missing the 5G wave by two years or more. Apple has made astute use of trailing-edge technology before (for cost reduction, while featuring leading-edge technology in a marquee role). But being that far behind in 5G would be highly problematic at a time when smartphone sales are slowing, and competitors are looking for every edge. Why would anyone buy an expensive 4G iPhone when Oppo, a Chinese company, is offering a 5G phone for half as much? Ecosystem lock-in, perhaps, but there’s only so much trailing you can do before customers abandon ship. Thus, the fulcrum of Qualcomm’s leverage was this little piece of silicon. It’s sort of like the bile duct in the human body. Tiny part; can’t function without it. And there is no substitute.
And although Qualcomm’s stock shot up in the immediate aftermath, Apple’s didn’t do badly at all, as investors realized that the Cupertino colossus would also benefit by getting a 5G phone to market faster. It’s also true that in the final hours of trading last Tuesday, when Qualcomm shares rose 23%, Intel’s stock fell in a nearly symmetrical way, perhaps heralding the announcement a few hours later that the Santa Clara-based silicon maker would exit the 5G modem business entirely. There is a lot of speculation that Apple may absorb Intel’s modem assets one way or the other in the next short while as the phone maker continues to seek a way to control all vital components.
But why did it have to come to this? An expert witness in technology cases (not this one) told me table stakes in a patent dispute start at $10 million. Clearly, both parties had the doh to take this dispute as far as it could go, but it was at base a commercial disagreement, a simple contract negotiation that had turned into a highfalutin antitrust case.
The main case was Apple’s against Qualcomm, and it had been consolidated with cases brought by Apple’s contract manufacturers, which are the entities that actually paid Qualcomm. In the jerry rig that existed before the settlement, Apple actually had no license with Qualcomm. Apple’s rights to the technology were embodied in licenses held by the contract manufacturers (Compal, Foxconn, Pegatron, and Wistron). When Apple decided — most cynically, it turns out — to go after Qualcomm, it told these manufacturers to stop paying Qualcomm because, if they didn’t, Apple would not compensate them, as had been the norm, and they would simply be out the money.
During the two-and-a-half-hour window that gave us a view into the parties’ thinking, Apple’s lead attorney, Ruffin Cordell, painted a picture of Qualcomm as a monopolist run wild. He said it was “wrong to force people to pay twice” (once for the modem and once for access to the intellectual property) and used a metaphor to illustrate. It was as if, when consumers went into Kentucky Fried Chicken and ordered a bucket of chicken, they are asked to “pay extra for the Colonel’s special sauce.” He made repeated references to having to pay an “eating fee,” “double-dipping,” the onerous “Qualcomm tax,” and “no chips, no license,” the latter referring to an unfounded allegation that Qualcomm would refuse to license its patents to any manufacturer that didn’t have a portfolio license. He said, “They want an eating fee, even if I go to Popeye’s.”
About the license fees, he said, “They charge way, way, way too much.”
In a dismissive tone, Cordell talked about the uselessness of many of the 130,000 patents for which Apple was being required to pay, citing a patent for a “shirt and pants that have wires running through them.”
Perhaps most ironically, Cordell closed with this: “If you make good products, people keep coming back.” Of course, he was referring to the iPhone, but he could just as well have been taking about Qualcomm’s 5G modem, for which Apple eventually came back.
Next up was Richard Doren, representing the contract manufacturers. His argumentation largely mirrored that of Cordell. Until the Apple case, the manufacturers had been apparently happy to pay their license fees, as they were made whole by their customers. But in this context, they became part of a Greek chorus wailing and moaning about Qualcomm’s behavior.
If I were a juror and arguments had stopped at that moment, I would have been inclined to find for Apple. Cordell and Doren had painted a broad picture of a monopolist out of control.
Even after Evan Chesler, Qualcomm’s lead attorney, began his spiel, things weren’t looking that good. He spoke more softly than the other lawyers and seemed to hesitate at times. But he warmed up, and pretty soon he was in full throat. He began to show the impossibility of Apple’s contentions. For example, he told of how Irwin Jacobs, the founder of Qualcomm, had managed to get Motorola and AT&T to agree to take licenses at the 5% rate when Qualcomm had no leverage at all in the industry. “All Jacobs had was an idea and a van,” he said, referring to the mobile demonstration unit Jacobs used to show that his cellular technology was superior to what was in the market at the time.
Perhaps Chesler’s finest moment was when he refuted the “useless patent” allegation. “They come to court and make fun of us for patenting pants!” he bristled, and then described how the so-called pants patent was part of a family of global positioning system (GPS) patents that included a possible application for finding children and pets by embedding GPS in clothing items.
He upended the fried chicken metaphor by noting that chips and licensing are two different things: “You pay for a chip when you buy it. When you buy a license, you get all the [related] technology. It’s more like you bought chicken and potatoes.”
To rebut the accusation of charging “way, way, way too much,” He compared the actual fees Apple paid to various Apple products like a proprietary Apple charging cable, a charging dock, a phone case, and a pair of AirPods, noting “$13 is the cap” for licensing fees, and Apple actually paid less.
So, Qualcomm’s case was looking stronger and stronger minutes before the settlement was announced.
But before going out for celebratory drinks, we should take a closer look at the facts that Chesler was able get into the record. He was able to show, via internal documents that came out during discovery, that Apple forced the contract manufacturers stop paying Qualcomm (by saying that any payments made would not be compensated), demanded that they not settle, make any admissions of liability, or do anything to “harm Apple’s position.”
He demonstrated, again through internal documents, that Apple valued Qualcomm’s patents more highly than other technology suppliers like Nokia, but nonetheless created a program that attempted to devalue them by “reshaping FRAND.” FRAND, which stands for fair, reasonable, and non-discriminatory, is the international regime under which Qualcomm’s most important patents are licensed. This reshaping would be done by reducing the base for royalty calculation from the whole phone to just the modem, knocking the royalty rate down, excluding various other elements, and “building favorable, arms-length ‘comp’ licenses.”
This last took a degree of cynicism difficult to fathom. What Apple was saying in effect was “Let’s use our purchasing power to cut some aggressive deals with other technology suppliers at lower rates, selectively filter our deal pipeline to play up these deals, and then point to them as illustrative of how Qualcomm is overcharging.”
While Apple was publicly dissing the value of Qualcomm’s portfolio, privately it said, “Compared to others, Qualcomm holds a stronger position … [in] cellular and WiFi [patents].”
Documents also show that, well before it initiated legal action, Apple knew its plan amounted to tortious interference (interfering with a contract to which it was not a party; i.e., between Qualcomm and the contract manufacturers), putting Apple “at risk for infringement, tortious interference, and full royalties (plus any interest, penalties, etc.).
Apple planned to introduce Intel chips to its mix two years before actually doing so in 2016, even though its contract with Qualcomm (for business cooperation) specified certain benefits to Apple (including substantial payments) that would occur only if Qualcomm remained a sole supplier. Again, super-cynically, Apple waited to sue Qualcomm until after Qualcomm had paid $4 billion to Apple under its agreements.
And there are more such manipulations too numerous to include here. But in sum, they demonstrate how Apple hoped to use its bulk and market power to cripple Qualcomm’s business. It stopped paying royalties to Qualcomm even as prices and profits for its own products rose.
The business reasons for Apple to undertake all this skullduggery is clear: smartphone market growth has slowed, Apple no longer has a multi-year technological lead on its rivals, and the only way to please a highly exigent Wall Street is to squeeze more profit out of this slowing revenue growth. Increasing profit when revenue growth ceases can be done only by lowering costs, and Apple has a history of pressuring its suppliers mercilessly. From Apple’s point of view, Qualcomm was just the best squeeze within reach.
Given all the bad blood, where to from here? The settlement involved a large payment from Apple to Qualcomm, representing unpaid back royalties, a six-year direct license between the parties (with a two-year possible extension), a multi-year chip-supply agreement, and a withdrawal of all litigation worldwide. Qualcomm has indicated that the agreement will add $2 per share in earnings “as product shipments ramp.”
So, all good, right?
Not quite. There remains the matter of the Federal Trade Commission (FTC) suit, which is still awaiting a decision. The FTC’s case was always weak, but U.S. District Judge Lucy Koh of the Northern District of California is no friend of Qualcomm’s and could still decide against the company. That outcome could yet upend Qualcomm’s licensing business. My own bet is that Judge Koh does not want to render a decision that would destroy the U.S. patent system in her court, nor sustain the career damage that would occur if her decision were overturned (highly likely) on appeal.
Then, there is the question of whether Apple will take another bite of the Qualcomm at a later point. Apple is still Apple and will continue to seek ways to lower costs, gain leverage over suppliers, and maintain control over its vast empire. It has opened an office in San Diego (Qualcomm’s headquarters city), hired away some Qualcomm engineers, and clearly seeks to bring the cellular modem in house, as it has already done with the main processor and graphics unit. With the settlement, Qualcomm’s trade-secret-appropriation charge goes away, and Apple has the source code for 5G modem software. The leopard may have taken a bath, but its spots are as distinct as ever.
The slides that made it into the public record also provide an avenue for the federal government to look into Apple’s practices in general. We in the industry have known how Apple rolls for years, but the company has cleverly managed to maintain its image as the underdog, even as it has come to dominate the technology industry. For example, in this suit, Apple tried to make Qualcomm look like the big bad wolf when Apple is 10 times Qualcomm’s size. The slides, which show how Apple wanted to hurt Qualcomm financially, are like the Nixon Tapes, a smoking gun of bad behavior. The Department of Justice or the FTC might want to look further into what the evidence revealed.
In the end, the quality of the technology may be the governing factor. Qualcomm has always contended that its customers buy its products because they are the best, not because buyers are coerced. Yes, Apple may try to make a 5G modem with expertise snagged underhandedly from Qualcomm. But with each generation, the difficulty of the problem increases. Right now, Apple’s rivals Huawei and Samsung have integrated their 5G modem into their overall processing chip, a so-called “system on a chip” (SoC), giving them advantages in cost, power, and performance. Apple can’t do this with someone else’s modem and its own processor.
The just-signed agreement gives Apple up to eight years of 5G modem supply from Qualcomm, which seems about right because it takes about that long to develop a modem. But there’s another piece of the communications system — the radio frequency (RF) front end, which covers the area between the antenna and the modem — that Apple cannot easily address. In 5G, the RF front end is more tightly integrated with the modem than before. And that integration extends to the power amplifier as well. The RF front end reads information going over the air and feeds it back to the other components so that they can take advantage of knowing what’s coming to adjust their behavior. This integration is critical for keeping performance high and power consumption low, particularly when 5G channels will be as much as 10 times wider than today’s 4G channels.
Apple will continue be at a disadvantage to the Android ecosystem with respect to power consumption because it cannot achieve full integration of all these components anytime soon. As one Qualcomm engineer put it, “the modem is an NRE (non-recurring engineering) money pit. Modems iterate faster than processors. You need a new one every year or two.”
One of Qualcomm’s in-house attorneys called the deal “one of the most significant settlements in litigation history, the first time a company stood up successfully to Apple.” There was some speculation at Qualcomm headquarters the day after the settlement that Apple’s premeditated attack on Qualcomm would one day be studied in a Harvard Law School case as “the most flagrant and egregious example of patent holdout that has ever occurred.” It just goes to show you, infringement is efficient — until it’s not.
But although Qualcomm won big, Apple didn’t really lose. What’s most important is that the U.S. cellular industry now has a way forward, and our great technology companies can work together in the competition with the Chinese for dominance in the 5G market. For that, we’ll need all hands on deck.