The Tech Industry Death Match That Wasn't
Mike Feibus
President and Principal Analyst, FeibusTech. Also Columnist at USA Today Tech, MarketWatch, CIO Magazine
WILMINGTON, DE – After a third full day of testimony here Wednesday in this infuriatingly persistent Arm v Qualcomm legal battle, US District Judge Maryellen Noreika freed the jury, sending the eight members out to face the crisp, sunny day turned damp.
And after the courtroom door closed behind the five women and three men, lawyers for both sides crowded the podium to begin structuring the next morning’s jury instructions.
Immediately, I sensed that Judge Noreika was thinking what I was thinking: “I don’t know what this is!” she told Arm’s attorneys. “What contract provision is in dispute?”
Yes, exactly.
The jury will resume deliberations this morning, having failed to reach a verdict after three-and-a-half hours Thursday afternoon.
I felt their pain. Under our belts were three-and-a-half days of sound and fury – tabloid-quality he-said, she-said testimony that alternately framed one side as wrongdoer and wronged. And I’m not feeling any more informed or enlightened about the relative strength of the cases of Arm or Qualcomm, two successful, longtime industry partners that drifted apart. Or of the path forward for their symbi-soured relationship. Or for the technology business overall.
Indeed, this case has felt about as relevant-adjacent to the electronics industry as a Mike Tyson-v-influencer bout is to boxing. But make no mistake, there are critical new business insights – all with serious implications – for any company, large or small, that is:
Because at the core of this court battle is concerning new behavior out of Arm, a publicly-held, increasingly for-profit gatekeeper for a de facto industry standard that guides how billions of chips interact with software, systems and each other.
Consider yourself warned
Over the past decade, Arm and Qualcomm have evolved into the worst best partners. Qualcomm was Arm’s biggest customer for a time, and they both made each other a lot of money. But they each wanted more.
For its part, Qualcomm wanted industry-leading processor performance, and to pay less on a per-unit basis for it. And increasingly profit-minded Arm wanted to build a bigger chunk of systems – and get paid commensurate with that. Regardless of how well their products performed.
Each companies’ desires grew more urgent in the 2010s. Apple, Qualcomm’s biggest smartphone processor rival, started making its own chips under an Architecture License Agreement, or ALA, with Arm. Qualcomm was paying more for pre-designed processor cores under an Arm Technology License Agreement, or TLA.
Arm’s profit priorities intensified after Softbank bought the company in July 2016. And the tension between the competing goals of Arm and Qualcomm came to a head.
Softbank thought it had an answer to its Arm profit problems in September 2020, when it announced that Nvidia was buying the company for a heart-stopping $40 billion. Qualcomm and others opposed the deal. And regulators listened, convincing Nvidia to drop its bid in February 2022. Later that year, Softbank pivoted toward taking Arm public.
For its part, Qualcomm completed its acquisition of Nuvia early in March 2021. By that time, Nvidia’s proposed buyout of Arm was already in regulatory crosshairs. And Qualcomm was working toward realizing its dream of supplying processors that beat Apple’s best – and paying Arm much less on a unit basis.
IP UH-OH
When Arm sued Qualcomm for breach of contract in August 2022, few observers – me included – ever thought it would go to trial. For one reason, Arm’s case was weak. Nuvia had an ALA in place if it ever needed one to build and sell Arm-compatible processors. But it didn’t need one, as it happened.
Qualcomm also has an ALA – a much better, more comprehensive license – with favorable pricing. So after the acquisition, the Nuvia team began tailoring its core for the laptop PC market. And in June 2024, the chips began selling under Qualcomm’s ALA.
That’s a pretty defensible design-and-go-to-market plan. That said, it would have been even more defensible had Nuvia founder Gerard Williams not insisted on adding terminology like design “derivatives,” “included technology” and “compliant cores” to its Arm ALA. Williams thought the language would add an extra layer of IP protection. As it turned out, though, creative Arm folks twisted their meaning in effort to protect their interests over Nuvia’s.
What damages?
Although Arm alleged breach of the Nuvia ALA, the IP licensing house never claimed any damages. A curious omission, given that assessing harm is a foundational tenet of contract law. That could be because Arm doesn’t want pocket money if it means allowing Qualcomm to continue shipping its own Nuvia-based processors, which earn Arm pennies apiece in royalties – and outperform Arm’s own off-the-shelf cores.
In any event, the jury isn’t considering damages. They’re not considering much at all.
Regardless, the potential ramifications of the jury’s decision are gargantuan – for both Arm and Qualcomm as well as for the entire electronics industry. That said, the most likely outcome will be that the two companies continue operating more or less as before, with some minor adjustments.
In the meantime, though, this case should put the industry on notice. The testimony in this lawsuit has laid bare Arm’s intent to leverage its gateway status to maximize profits – even if it means putting a big partner at risk.
In the short term at least, partners – partnenemies? – need access to the Arm instruction set in order to link their products to a globe full of smartphones, webcams, thermostats, kitchen appliances, automobiles and laptops – along with the applications that make those systems useful.
But they’d better start working together on a long-run alternative. Because Arm’s profit squeeze is only just getting started.