Qualcomm vs Apple – Black Magic

Apple won’t be able to pick a fight in 2026.

• As I have long suspected, Apple is really struggling to make a high-quality 5G radio chip meaning that Qualcomm could end up supplying Apple with radios well beyond 2026 when the current contract expires.
• This will also make the patent renegotiation with Apple much easier as the current agreement to pay Qualcomm royalties on cellular devices also expires in 2026.
• Bloomberg (see here) has spoken to a number of “people with knowledge of the project” (as I have) and learned that Apple is further behind than even I had been led to believe.
• The general assumption on the street is that because Apple is arguably the best processor maker in the world, it should have no trouble replicating the chips that control cellular communication.
• Anyone who has been around for a while (like me) understands that when it comes to cellular radios simply being good at designing chips is not nearly enough.
• There are three main problems:
  • First, Analogue: Data travels through the air as oscillating radio waves and the information is not digital but analogue.
  • This is like the difference between music that is recorded on a vinyl disc (analogue) and music on a compact disc (digital).
  • The analogue data is received by the antenna and then converted into digital information which the device can then use.
  • To make a working radio system, one has to have a very good ability to execute in both the analogue and the digital domains.
  • The problem with analogue is that understanding the laws of physics is not enough and in many cases, experience in analogue is just as important as engineering knowledge.
  • This is why analogue radio is often referred to as a black art explaining why so many companies have completely failed to get to grips with this part of the radio.
  • This is where Apple has come unstuck as the business that it bought from Intel which originally came from Infineon had completely failed to make a working 5G radio and had also had great difficulty with 4G.
  • Very little progress appears to have been made meaning that Apple remains years away from being able to make a 5G radio that can match the performance of the one it buys from Qualcomm.
  • Second, Radio Frequencies: The amount of radio spectrum available for cellular communications is quite limited.
  • If the frequency is too long, you can’t get enough data into the radio wave and if it is too short, the radio wave won’t travel far enough to be able to create an economically viable network.
  • This range is roughly 450Mhz – 4Ghz but even at 4Ghz, things are already getting a bit dicey on the distance the radio wave can travel.
  • One of the appeals of 5G is that it is supposed to be able to deal with frequencies above 6Ghz (mmWave) which adds another layer of difficulty in terms of making the radio work correctly.
  • However, the real problem is that the number of radio bands that a radio chip that works anywhere has to support is growing all of the time.
  • It is fiendishly difficult to be able to simultaneously support all of these bands at the same time and do so in a small and power-efficient package.
  • This and mmWave are two areas where Qualcomm really excels and Apple is years away from being able to catch up.
  • Third, moving target: Radio technologies are constantly evolving and as a major contributor to the standards, Qualcomm has advance knowledge of what is likely to be included in the next releases of the standard.
  • This is one factor that explains how Qualcomm has been able to come up with working chipsets for a new radio standard long before anyone else can.
  • It has repeated this feat for every cellular radio standard since CDMA in the 1990s with the exception of GSM to which it did not contribute.
  • This means that Apple has to both catch up with where the technology is today and ensure that it has high-quality working radios available for its devices along the same timeline as Qualcomm.
  • Apple has to both run very fast and hit a moving target at the same time.
• The net result is that I don’t think Apple will be close to being able to replace Qualcomm by 2026 meaning that it will be forced to extend its contract again.
• These sorts of patent licenses typically last for 10 years but when Apple signed with Qualcomm in 2019 (see here), it only signed up for 6 years.
• It was its requirement for a working 5G modem that forced it to back down in its fight with Qualcomm over patent royalties and sign up for a new license.
• I have long suspected that Apple signed up for 6 years because this is how long it thought it would take for it to create a working 5G modem so that it could ditch Qualcomm and resume hostilities over its royalty rate.
• This plan looks like it is coming unstuck because I continue to think that Apple will not have an internal working 5G modem by April 2026 meaning that it will be in the same position as it was in 2019.
• This is excellent news for Qualcomm as its internal assumptions for how much share it will have at Apple (20% in FY 2026) are clearly too low and could easily still be sitting at 100% when that time comes.
• Most importantly of all when it comes to patent license negotiation time, Apple will not be in a position to pick a fight and so it will have to grit its teeth and at least extend the contract for 2 years until 2028.
• In 2028, everyone may be talking about 6G and then the whole cycle will repeat itself yet again even in the unlikely event that Apple cracks 5G by then.
• Hence, I think Qualcomm is in a pretty safe position for at least the next 5 years making the shares look increasingly attractive as the valuation remains pretty low and the risks are declining.
• I have a position in Qualcomm and remain quite happy to sit on it.