Disaggregation And Re-aggregation

Big changes are afoot, but it’s still hard to figure out what the final landscape will look like.

popularity

The proliferation of platforms, subsystems and IP of any sort, as well as the move to stack die in 2.5D configurations, will force a realignment of the ecosystem.

For the moment, it appears that vertically integrated companies such as Apple and Samsung have a distinct advantage. It remains to be seen just how substantial that advantage really is, however. As chips become a collection of more merchant parts that can be assembled quickly from a menu, things will change significantly.

The question isn’t so much whether the semiconductor industry disaggregates or re-aggregates. That’s the natural rhythm of business, particularly a technology business that is constantly evolving and spiced with new inventions. The big question is how the pieces come together and eventually pull apart, and what gets pushed aside or re-formed along the way.

Stacked die are the most intriguing part of this development, and one of the key drivers in the changing landscape. Being able to assemble a collection of commercially available platforms, and custom IP for specific markets—sensors, MEMS chips and analog parts—will force fragmentation in some markets and consolidation in others.

The need for platforms constructed of bleeding-edge digital logic and memory will create huge markets for some companies. In addition, the ability to create customized solutions for market segments will create small but profitable markets for others. The ones that will be left in the lurch are the companies that try to create specific solutions for broad markets, or general solutions for small ones.

This is an important shift. It explains why companies are working with Google to improve the power consumption and performance of Android, and it explains the thinking behind the rumored Facebook phone. It also explains why investors are becoming skittish about Apple’s future, and why Intel has been making huge investments in places it wouldn’t ordinarily invest.

Put in perspective, billions of dollars already have traded hands and we still don’t have a clear picture of what the final landscape will look like. Were these just opportunistic investments, or were they made for a bigger purpose? Our bet is the latter, given the number of changes that have been made recently and the seemingly strange places that money is spent. Given all of this activity, things are almost certain to look different in several years, but how different is still a mystery.

—Ed Sperling