Last week, Qualcomm filed its second response to Arm’s suit, and it makes for some eye-opening reading. In addition to some fairly strong-worded denials of all of Arm’s claims, in the Factual Background section towards the end of the filing Qualcomm makes the following claim:
- ARM also told one or more Qualcomm customers that, when the existing TLA
agreements expire, ARM will cease licensing CPUs to all semiconductor companies—including
Qualcomm—under an ARM TLA. ARM claimed that it is changing its business model and will
only provide licenses to the device-makers themselves. ARM has explained to the OEMs that a
direct OEM license will be the only way for device-makers to get access to ARM-compliant chips.
In plainer language, Qualcomm claims that Arm is telling companies that it plans to change its licensing model, that it will move from licensing chip designers to licensing chip designers customers instead. What?!?
If true, this marks a major shift for Arm, and as Dylan Patel pointed out one that is likely to be very hard for Arm to achieve.
What is going on here?
It is very important to remember the source of this claim. Qualcomm’s lawyers are in take-no-prisoners mode, pushing back very hard against Arm’s suit. While they are not going to fabricate anything, they are going to paint anything Arm does in the harshest possible light. An off-hand comment from a low-level engineer musing about some brainstorming session could possibly be the source. Probably not, but let this serve as a reminder to think of every e-mail as viewed by opposing counsel in some future lawsuit before hitting send.
Lawsuits like this can be thought of as giant games of chicken – the loser is the one who blinks first, regardless of the specific merits of the case. As such, when big corporations fight, they are very keen to apply pressure in channels beyond the courtroom. The Apple v. Qualcomm suit is a good example. Apple was ‘happy’ to spend a few billion dollars in legal fees, but ultimately caved when Intel’s 5G modem got delayed (again) leaving Apple with no choice but to settle with Qualcomm so as to gain access to the 5G modem which Apple really needed. In this suit, Qualcomm presumably would view it favorably if every other Arm licensee called Arm in near-panic for clarification about their access to future Arm designs. And there are other reasons to doubt this claim. The language in this section is markedly vague in a few places, with lots of pronouns where it is unclear to whom they are referring. And is it CPUs that are not going to be offered under this model or all chips? There is a vagueness in this material that stands in contrast to much of the rest of the document, which is incredibly precise in the way that extremely expensive attorneys do so well.
So we tend to treat this claim buried deep in an 83-page document as just a bit of noise or scare-mongering. Arm is not really going to change its business model. Is it?
Changing business models is not easy, very few companies have ever done it successfully. True, start-ups pride themselves on their agility and ability to pivot, but that is something that is possible for a company with $0 in revenue and 20 employees, not something for a company with billions in revenue and 6,000 or so employees.
Imagine the scenario where a non-technical company wants to buy chips. Does Arm have salesman contacting them today? Who do they speak to? Does the procurement team there know who Arm is or what they do? Has the CFO ever heard of them? Some engineer in a non-core team has to get approval from a dozen managers just to enter into a licensing conversation. Approval would require input from finance, legal, engineering, operations and probably the CEO. No one likes those kinds of cross-functional conference calls, and they will like it even less when they hear that they have to sign this agreement or face losing access to their chips. Months of MS Team meetings where the first 15 minutes of every call has to start with an explanation of what Arm does and who they are.
And how does that conversation go? If the OEM wants chips, they have to sign a complex Intellectual Property (IP) agreement, pay a big upfront license payment and an ongoing royalty. That just looks like a new expense, another thing everyone hates. Arm could argue that they are already paying this indirectly through their chip vendor, which is fair. But will that chip company offer a discount to reflect this new expense? Which leads to a whole other series of angry conference calls.
The time element is important here. We once gave a one hour talk to a Fortune 1,000 company about wireless topics. When we tried to collect our speaking fee, a whole $500, we had to spend three months navigating the company’s Accounts Payable team and all their 1990’s software portals. It will take years for these companies to bring on a new vendor like Arm. Qualcomm’s document says Arm plans to implement this new model in 2025, which is very soon when viewed in this context.
Surely, it cannot be that hard, after all Qualcomm runs exactly this model charging OEMs a royalty for the Qualcomm IP in their hardware. True, but Qualcomm was only able to achieve that through a few decades of bare-knuckles legal brawling including a half dozen billion dollar+ lawsuits, a half dozen anti-trust investigations around the world and 20+ years of intense, sharp-elbowed ‘negotiations’ in the wireless standards bodies.
Does Arm really want to go through all of that? Do they really think it is possible to transition to this model while they are preparing to go public? It is much more likely that Qualcomm’s attorneys are building an Everest out of a mole hill as a way to apply pressure through other means. In our first review of the lawsuit, we cautioned that when Arm filed this suit they would have to face some dirty laundry being shared in the public. This is likely to be just the first salvo of that onslaught. As we said above, we do not think Arm is really going down this path, but if they do it will be a good time to find a way to invest in RISC V companies.