Charged with answering three questions, the Arm versus Qualcomm jury returned a verdict on only two: whether Qualcomm breached Nuvia’s Arm licensing agreement (ALA) and whether the Qualcomm ALA covers the acquired Nuvia design. The jury did not rule whether Nuvia breached its ALA. Therefore, Arm argues the result is a complete mistrial and has stated it intends to seek a retrial. The judge directed the parties to seek a mediated solution instead. It’s hard to believe that they can agree on anything, but the judge made it clear that she doesn’t want another trial.
My Three Questions
Although both parties delivered long and winding arguments, the jury ultimately only had to answer three questions:
- Did Arm prove by a preponderance of the evidence that Nuvia breached Section 15.1 (a) of the Nuvia ALA?
- Did Arm prove by a preponderance of the evidence that Qualcomm breached Section 15.1 (a) of the Nuvia ALA?
- Did Qualcomm prove by a preponderance of the evidence that the Qualcomm CPUs that include designs acquired in the Nuvia acquisition are licensed under the Qualcomm ALA?
Section 15.1 (a) stipulates that upon its ALA’s termination, Nuvia must discontinue using the technology it developed. Implicit among these three questions is another one: Can the jury infer that Qualcomm assumed the Nuvia ALA? Moreover, the jury also had to assess whether Qualcomm or Nuvia harmed Arm, a necessary condition to find a breach.
The jury’s job, in summary, was to interpret the company’s behaviors and contracts—appropriate tasks for eight regular people. Interpreting the contracts entailed evaluating them as a whole (not by considering isolated parts), looking at their language, and considering how the parties acted before any disagreement arose.
Questions 3 and 2
Reviewing the questions in reverse order, the Qualcomm ALA had language that Arm said excluded the acquired Nuvia cores. Specifically, it covered CPUs Qualcomm made or had made. Narrowly interpreted, this phrasing would exclude cores obtained through acquisition. A common-sense interpretation, however, would be that the ALA covered any cores. Moreover, initial Arm internal communication showed that the company held this viewpoint, too, changing its stance only after the chairman and CEO decided to pursue Qualcomm. Thus, although Arm argued its case clearly, the jury didn’t take the company’s side.
With its ALA covering the Nuvia cores and the question of whether it assumed Nuvia’s obligations, Qualcomm could hardly be seen as breaching a contract to which it wasn’t a party. During the trial, Qualcomm insisted that it didn’t assume the contract and highlighted that Arm was not seeking damages—and, therefore, suffered no harm.
Question 1
The first question depends on whether Nuvia’s design, even stripped of Arm-specific features, is Arm’s technology. Here, too, Arm pointed to a specific ALA clause stating that it is. Other clauses differed, such as by indicating that the design isn’t Arm’s but instead follows from it. Qualcomm, however, didn’t highlight these clauses, which would’ve aided the jury in interpreting the ALA as a whole.
Moreover, Qualcomm wasn’t effective in arguing that architecture and microarchitecture are separate (or at least separable). The company stated it removed and rebuilt some aspects of Nuvia’s design but was neither clear nor convincing. Nor did the company rebut Arm’s ridiculous claim that because the Arm v8 architecture specifies the add opcode, the Nuvia CPU’s adder is Arm’s technology.
Qualcomm’s unconvincing arguments and Arm’s inadequate delineation of which microarchitecture features tie specifically to its technology led the jury to hang on Question One. Having implicitly found Qualcomm did not assume Nuvia obligations and the company, therefore, no longer exists, the question is moot. For now, it’s the last straw for Arm to grasp.
Bottom Line
The trial shed negative light on both parties, but Arm suffered more reputational damage. Having long jealously guarded its intellectual property and aggressively promoted its wireless technology, Qualcomm’s antagonism is well known. By contrast, Arm was a collegial, plucky little British company selling CPU designs to chipmakers worldwide. Now, we see it’s at least as antagonistic as Qualcomm and greedier.
In pursuing this case, Arm never saw licensees stampeding to RISC-V as a likely outcome if it prevailed. The upstart trails the Arm architecture in application processing. But, with the billions of dollars at stake now quantified, all can see the business case for getting RISC-V up to snuff. Assuming the partial verdict holds, the ruling in Qualcomm’s favor staves off a stampede. But even as Arm expands its market with PC- and server-processor share gains, its ambitions, laid bare by the trial, set the stage for its demise.