Sic Transit Gloria Mundi?

A recent comment in one of our earlier posts raised an interesting train of thought that we wanted to share:

I am astounded by the quality and writing that you all consistently put out. Thank you. Surprised that this is free. But do keep it so!

In reply to: Slab Bets the House on IoT

Wait, sorry not that, one. This one:

I won’t be so sure about Apple (or any other company), the demise takes only one failed generation.

In reply to : The Handset Industry is A Flat Circle

The commenter’s point is that Apple may look like it is dominant now, but that strength is fleeting. All it will take to topple them is one bad generation of product. Our first instinct was to agree with this. The history of consumer electronics demonstrates this clearly. For instance, Motorola’s stock used to rise and fall entirely on their ability (or lack thereof) to ship their latest clamshell phone in time for the Q4 Holiday Shopping Season. Nokia went from industry leader to out of the business in three years because they could not make the turn to smartphones. This was also common wisdom in the semiconductor space as well.

But it no longer holds. The difference can probably be summed up simply as ‘software’. We are all now tied pretty tightly to our smartphone software. Porting data to a new device is hard enough even on the same operating system (OS) to make switching to a new OS prohibitively painful. Then factor in all the Apple services like Music and TV+ which add to the stickiness of the platform.

Another important factor is that replacement cycles have extended considerably. In the feature phone era the average consumer bought a new phone roughly every two years, but now those replacement cycles are closer three to four years (which makes sense when phones cost $1,000).

Adding this together, it seems unlikely that a single bad product cycle could kill Apple. What would the worst case scenario look like? Let’s say they design the Homer Phone, a phone so bad that no one buys one that year. The company has $204 billion in cash on its books now. Their operating expenses are about $40 billion a year, capex is about $20 billion. They would also take a big loss on phones purchased but not sold, so 50 million units at $600 a piece, another $30 billion. Add it all up and the company could survive two years on just their cash on hand, to say nothing of their other products. It would be ugly, but it is also unlikely.

In all fairness, it is worth asking this question. As Apple becomes more vertically integrated, they become more vulnerable to their own shortcomings. We think of this a lot when we look at Apple Silicon. Today, this part of the company is executing incredibly well, they are the world’s best run semiconductor company, and they don’t sell chips. But the risk here is what happens if that execution ever slips. The timelines for semis are incredibly tight, and hardware cannot be fixed once it ships. So what happens if one part of the chip design is not completed on time, say some new, advanced AI core, that then causes the whole A Series or M Series processor to be delayed which delays the shipment of the phone or laptop it powers. These are real risks, which though unlikely today at Apple have still plagued every other chip company ever. The other, probably more significant risk, is to look at it from a different direction. Apple has to be cognizant of these risk and has likely factored this into their planning. So does that mean their risk mitigation measures cause them to be more conservative in their chip design? Are they less willing to take chances because everything is so tightly integrated? Could this lead them to miss a cycle? Probably not, but definitely worth considering.

That being said, as things stand today, Apple looks fairly unassailable. The glory of this world always fades, but that does not mean it will fade any time soon.

Photo by Simpsons Fandom

2 responses to “Sic Transit Gloria Mundi?

  1. Just curious, but have you heard of Rain and it’s APUs? Wondering how you think this impacts NVIDIA (and also to some extent Apple now that they’re designing their own chips). https://rain.ai/technology

    On Tue, May 4, 2021 at 1:08 PM Digits to Dollars wrote:

    > D/D Advisors posted: ” A recent comment in one of our earlier posts raised > an interesting train of thought that we wanted to share: I am astounded by > the quality and writing that you all consistently put out. Thank you. > Surprised that this is free. But do keep it so! In re” >

    • I didn’t know Rain, but am aware of a few start-ups doing something similar, AI in analog, interesting approach, might work for some applications, but has some limitations too.

Leave a Reply