Acquisitions and the leading edge of design defined 2015, but more changes are ahead.
As tech journalists we are devout interpreters and re-interpreters of our own statistics. We can track how many clicks a story gets, which region or country those clicks come from, and how much interest there is in various subjects over time.
The publishing industry always has been driven by data. In the early part of the 20th century, newspapers were engaged in circulation battles, which were the only statistics available at that time. In cities with more than one newspaper, the circulation leaders (or those that were closing in on the leaders) would post their numbers at the top of the front page. And they would hide them from readers when they were trailing and sales were slumping.
Online, statistics are instantaneous and potentially much more useful. It’s true that they can be misleading—a story on one subject might draw huge traffic one day, but a similar story might draw far less traffic the following week. Still, those inconsistencies tend to disappear over time and in context. And in our slice of the world, which is semiconductors, some stories will never pull in huge numbers even if we have reached every reader on the planet who cares about a particular subject.
But overall, the biggest stories of 2015 tended to fit rather neatly into two buckets—consolidation, and problems in moving to the next process node. Those are clearly the two biggest trends of the year for the semiconductor industry, and there are some arguments that they’re different sides of the same story. While consolidation has been a persistent theme in this market for some time, in past years it involved far smaller companies with new ones refreshing the market at equal or greater rates than acquisitions. That isn’t happening anymore, which means there are fewer companies to buy, the prices are higher, and the disruption is larger.
Big acquisitions involve thousands of people spread across sometimes dozens of countries and often as many product areas and verticals. They affect everything from macro capital markets to which technologies will be successful in the future. And as companies such as Apple move more technology development in-house, through a combination of acquisitions and organic hiring, those acquisitions take on new significance. It shouldn’t be a shock that this subject commanded such a strong following.
The second subject area is harder to wrap your head around because it doesn’t have any firm boundaries or numbers. The push to finFET-based process geometries is not an equivalent jump to moving from one node to the next in planar transistors. And the move from one node to the next in the finFET world is not proportionate to the previous generation of finFETs. While Moore’s Law from a technology standpoint is still achievable, from an economic standpoint it’s obviously no longer relevant for everyone. And while it may be simpler to pattern with EUV lithography, assuming it becomes available, no one is certain that will be cost effective because at this point it is unreliable.
These are the drivers behind exploration in 2.5D, fan-outs and 28/22nm FD-SOI, and it is reflected in the number of readers on those subjects. There are widespread concerns that affect every aspect of the supply chain if the bulk of the market shifts toward new architectures or materials at older nodes.
Issues such as the IoT/IoE, security and embedded software are growing in importance, and reader traffic on those stories is up significantly year-over-year, and even from the beginning of the year to the end of the year. Stories about China are gaining interest, as well, as the country’s investment strategy begins to play out.
Fundamental shifts such as these drive the need for more information, and the bigger the changes, the greater the number of people who are affected and the more they need new information. Along with that, it also becomes possible to identify new trends that will have far-reaching effects. While none of these are confined to a calendar year, growing interest and occasional spikes do represent real shifts in how the industry is thinking and what it is thinking about. And there is plenty of change ahead that will alter attitudes about what is important in 2016 and for many years to come.