What a day yesterday! The biggest news of the day, probably even the month, is that Apple and Qualcomm have settled their dispute over royalties and licensing and dropped all litigation. There is so much to unpack about this and the broad implications for the industry. But the day was not over! Hours after the news broke that Qualcomm and Apple settled, which ultimately led to a deal for Qualcomm to supply chips to Apple and Apple acquiring a license to Qualcomm’s patent portfolio, Intel announced they were excited the 5g smartphone modem business! The day felt like weeks when all was done, but each event was entirely linked. Understanding how we got here is the crucial first insight.
The Long Road Back to Qualcomm
Bear in mind, and I’m trying to connect the dots. But, I know quite a bit about what chatter goes through the supply chain, so I feel like my hunch is correct with my interpretation of this timeline. Ultimately, I think Apple’s end goal had always been to acquire a Qualcomm license on their terms, not Qualcomm’s. My reasoning? Apple was single-handedly keeping Intel’s smartphone modem business afloat, and doing most the heavy engineering for Apple, simply so they could have some leverage on Qualcomm. This was one step in a longer plan. Apple then organized some tactical lawsuits, including heavily lobbying the FTC and ITC regulatory bodies to file suits against Qualcomm on the basis of antitrust. Apple was spreading a wide net, hoping to have a few favorable rulings which would force Qualcomm to make some changes to their licensing model, or at the very least, give Apple even more leverage to negotiate an extremely favorable license deal.
Some people have theorized that Apple wanted to dismantle Qualcomm’s licensing business model completely. This theory has never set well with me, because if you play that out to its logical conclusion, you realize that if Qualcomm’s business model was significantly impacted to the point that they can’t charge what is fair for their patents and have to charge much less, then that is actually very good for Apple’s competition. So if this was Apple’s goal, they would actually be enabling their competition to access a rich technology portfolio at a much cheaper price and as a result, be even more aggressive with their smartphone prices against Apple’s. In fact, one could argue that disrupting Qualcomm’s licensing business model would be disruptive to Apple because it would enable a significant amount of innovation by hardware companies who can offer much lower prices than they do today.
Apple’s goal in these tactics was simply to gain more leverage on Qualcomm than they felt Qualcomm had on them. Part of this leverage would come if they won a few favorable lawsuits that devalued Qualcomm’s patent portfolio. Unfortunately, a few of those did not go their way. One, in particular, I found interesting, was a suit that Qualcomm brought on Apple around patent infringement for which a jury awarded a modest $31 million dollars to Qualcomm. But the most interesting part of that case was the jury valued just three Qualcomm patents at $1.41.
Qualcomm has over 80,000 patents, and while it would be impossible to set a price on each one, a price of $1.41 for only three could lead to an unwanted precedent by Apple.
Qualcomm’s defense in all of this was to defend its business model by bringing suits against Apple in patent infringement. Apple’s offense was to prove Qualcomm behaved in a monopolistic way. Qualcomm had won several other suits in Germany and in China were courts agreed Apple infringed on their patents. Quite simply, Qualcomm was successful in proving there was more value in their patent portfolio than Apple wanted to admit. Given Apple was seeking, and ultimately acquired via the settlement, a direct license to Qualcomm IP they needed to be able to justify the license + royalty + chip costs could not be justified, and they hoped the courts would help with that justification.
Bringing Intel into the fold as a dual source was another attempt at Apple to gain leverage over Qualcomm. During the days I attended the FTC trial against Qualcomm, it was interesting to hear Tony Blevins, Apple’s VP of Procurement talk about why they choose Intel and the bind they were in with Qualcomm feeling like Qualcomm could dictate terms to them, and they could do nothing. He also made the point to talk about how Apple always prefers to have multiple component sources because it allows them to negotiate better. So it’s clear Intel was there to add a second source and thus provide some leverage.
The only issue was Intel was not as nearly fully committed to smartphone modems as Apple wanted. From many of my sources, I heard multiple times that not only was Apple contributing vast resources to doing most of the engineering work on Intel’s modem but Apple also was the one who put forth the effort of getting Intel’s modem globally certified. Furthermore, Intel was taking a financial bath on the modem business and given the amount of work Apple was doing for Intel on the modem engineering, and I wouldn’t be surprised if Apple was paying next to nothing for the modem part.
Newly appointed CEO Bob Swan had also been confirming what I had known for years, that Intel was not interested in being in the modem business in the long-term. One of my game theories was always that Intel would ultimately exit the modem business and Apple would buy the IP to use in their own internal modem efforts.
When it comes to understanding the timeline that got us to this settlement, I believe Intel’s decision to leave modems was the first ball to drop. Apple’s ultimate decision was also likely swayed, if this rumor was true, that Intel was going to be late to 5G with a globally certified part. However, that was a symptom of the underlying problem of Intel not committed to the modem business and not investing in it as they should if they were serious.
Given how much work I know Apple was doing for Intel on modems, I think Apple knew for a while Intel was not a long term solution for them and I think that has been clear for 5G for over a year now. I am convinced Apple has known for a while they needed to go back to Qualcomm around the timing for 5G.
One thing I did not know that came out at the FTC trial was that Apple was never a direct licensee of Qualcomm. This is part of the reason they went on an offensive strike. Apple did not like the idea of having to sign a license to Qualcomm IP, pay for that license, then still pay for the chipset. They call this the no license no chips policy and their whole argument in court was this was illegal double dipping because Qualcomm was one business with two divisions QTL and QTC. One sells chips, and the other sells licenses, and they are deeply intertwined. Apple was skirting around this by using the Qualcomm license of its contract manufacturers like Foxconn, but they had several agreements they entered into with Qualcomm which included things like a rebate payment from Qualcomm to Apple, exclusivity of Apple to use Qualcomm chips, and some agreed upon payments and rates. But Apple did not have a direct license deal with Qualcomm.
As a part of this settlement, Apple has now entered into a direct license with Qualcomm that’s the part that interests me the most. The deal, as we understand it, is a six-year license to Qualcomm’s full patent portfolio and a multi-year chipset deal with some options going forward.
We do not know how much Apple has agreed to pay for a license, royalties, or for the chipsets but we do know, from the court trial information that came out that Apple has never paid more than $20 to Qualcomm and after the rebate payment deal their costs were $7.50. So while it is totally I guess I’d wager the total deal came out to somewhere between $5-7 per device.
The straight forward part of this deal is simply Apple will get 5g modems from Qualcomm and can use any patents they want and not have to worry about patent infringement litigation going forward. But the big question in my mind is how does this impact Apple’s own internal modem efforts. Here is where I have some theories to propose.
The first is the obvious one, and Apple can buy Intel’s smartphone modem IP and use that as a part of their own internal efforts. This would not conflict with the current Qualcomm deal since Apple’s own modem is not likely a feasible option until 2022 or later and would give 5g more time to mature as a technology and Apple the time to get the 5g part ready for the global landscape. The second theory is a bit more interesting but also feels less likely. Perhaps, and I emphasize perhaps, as a part of this direct license Apple has with Qualcomm, the two companies can work together on a solution that integrates Qualcomm’s modem onto Apple’s A series SoC. This is the ultimate goal to embed the modem onto the SoC as it brings in benefits of power and performance vs. using the modem as a separate part on the board. Also, as Apple gets into more small form factor wearable solutions, it is essential the modem be integrated into their SoC. That is why the end goal is for Apple to design the modem onto their chipset designs.
This seems less likely, but would be super interesting, because it has never been done before. The other factor is if that were to happen then Apple would be even more deeply tied to Qualcomm since it would take quite a joint effort to accomplish an Apple design SoC with a Qualcomm modem integrated on it. But, given how hard the modem business is, and how Qualcomm does have the best wireless technology, I do not think this scenario is out of the realm of possibility.
This has been a fascinating story to watch and follow, and there are still some important parts to address in more analysis of this relationship as we see how it plays out.