For the past 26 years, the same company has led the semiconductor industry in yearly revenue: Intel. While the Santa Clara firm makes a distinct minority of the total semiconductor silicon manufactured worldwide, it still leads the market in sales revenue — or at least, it did.
Samsung is now expected to surpass Intel in terms of quarterly revenue, thanks in part to huge upticks in demand for memory. Samsung expects to invest some $18.6 billion in expanding its South Korean facilities and boosting its memory and display manufacturing businesses.
“From the second quarter, Samsung will become No. 1 in market share due to the recent increase in data centers and demand for solid-state drives,” NH Investment & Securities analyst Peter Lee wrote in a note to clients.
Reuters also reports that Samsung’s profits are expected to jump 67 percent year-on-year, setting new highs. The Galaxy S8 is reportedly selling on-track with the Galaxy S7, with the Note 8 expected to launch this August.
An inevitable result
This shift may seem like a big deal, and of course, in one sense it is. But it’s also the inevitable result of a fundamental shift in which devices are driving the future of computing, and how consumer buying patterns have changed in recent years. When most of the profit (if not the volume) in the computing industry was in CPUs, Intel and other CPU manufacturers did extremely well. Data from IC Insights shows how the types of companies that dominate semiconductor sales have changed over the decades.
Not every name has changed, to be sure. But there are plenty of new entries and old companies that no longer make the cut. In this case, it’s not that Intel can’t grow revenue or unit shipments, as cloud servers and data center shipments increase over the next few years. It’s that even though the profits per-smartphone are much smaller than what Intel makes on a Xeon or a Core i7, companies like Samsung are finally shipping enough of them on their own silicon to make up the difference.
This is the future that Intel’s original vision of what we called “x86 Everywhere” was supposed to address, with enough x86 chips in mobile devices, tablets, televisions, and communication devices to make up the difference and keep Intel on top, both in total mobile and desktop shipments and system revenue. With Intel having now withdrawnfrom the tablet and smartphone market in favor of focusing more in data centers and the like, it’s position at the top of the semiconductor market is facing serious attack.
