There have been many acquisitions in the semiconductor industry lately, which naturally leads to investor speculation about what deals might happen next. Recently, The Wall Street Journal reported that Intel was in talks to acquire Altera, a major vendor of programmable logic devices.
Now there's a new rumor that Apple (NASDAQ:AAPL) might acquire British processor IP vendor ARM Holdings (NASDAQ:ARMH). This, in my view, is extremely unlikely -- and frankly, borderline nonsensical. Let's dig deeper.
What does Apple need with ARM?
ARM Holdings develops a wide range of semiconductor intellectual property, including processor instruction sets, CPU core designs, graphics processors, and physical IP. Its chips go into many different applications, the most prominent of which are mobile devices such as smartphones and tablets.
Apple is an ARM licensee; it licenses the latest ARM instruction sets and designs its own CPU cores around those instruction sets. Apple currently licenses graphics processors from ARM rival Imagination Technologies (NASDAQOTH:IGNMF).
Apple already gets what it needs from ARM. By the very nature of their business relationship, Apple can simply license any technology of interest that ARM develops. It's hard to see what Apple would get by shelling out the big bucks to acquire the company.
Apple would literally destroy ARM's value
ARM's revenue base comes from licensing intellectual property to many different semiconductor companies and collecting royalties from each chip its partners sell. ARM doesn't make much per chip, but because its partners ship so many units, those per-chip royalties add up.
If Apple acquired ARM, I find it difficult to believe Apple would do so with the intention of having ARM continue to develop and license intellectual property to the wide range of companies it currently serves, which include many of Apple's direct and indirect competitors.
This would mean that ARM's revenue stream, once under Apple's control, could dry up.
Furthermore, one of the ARM ecosystem's strengths is that the company's instruction sets are used widely. This means the industry is replete with engineers who understand ARM chip development and, perhaps more importantly, develop software and tools for the ARM instruction set. I believe taking away that scale by making it proprietary would defeat the very point of ARM's existence.
Apple could do what buying ARM would achieve organically
The one argument I could see for Apple buying ARM -- and this one is a stretch -- is to take complete control over the development of the ARM instruction set going forward. It's hard to see the business case for this, particularly given that Apple can (and likely does) collaborate with ARM to define future versions of that instruction set.
However, if Apple really wanted to take instruction set development in-house, it could conceivably develop one of its own and do a software transition from ARM to the proprietary Apple instruction set. Alternatively, Apple could spend a fraction of what it would on ARM and buy Imagination Technologies. Imagination owns the MIPS instruction set, which is a direct competitor to the ARM instruction set.