When you were young, you were the center of attention. Everyone wanted to be near you. Magazine covers. Fame. Attention. They named a whole Valley after you. But today, software is the hot young thing soaking up all the attention. You are still important. You provide the foundation on which everything else stands, a stable provider. But you move slower, the things you make are less shiny, the cool kids seem very young, you struggle to communicate with them, and the snacks you provide your employees are far inferior to what software doles out.
Such is the semiconductor industry today.
The industry was once white hot, but it has now grown and matured. The question that every company in the space has to answer is how do they conduct themselves now?
Many will just muddle along. Others are honing their retirement plans to produce large, but boring nest eggs. But a few of them are going still trying to recapture those glory days – buying a Harley and dating people far too young for them, positioning themselves as still hot.
The industry is generally profitable today, a big change from not so long ago. But it is growing much slower and there are no obvious catalysts to change that anywhere on the horizon. So companies are left with a range of responses. At one end of the spectrum is profit maximization and focus on shareholder value. At the other end, are companies largely living in denial, presenting themselves as offering hot new technologies in markets that are about to explode into hypergrowth, any day now. Most companies fall in the boring middle, struggling along in prosaic ways, enjoying healthy share prices and good conditions, but recognizing that these conditions, and the independence of their companies, cannot last forever.
The Godfather of the former end of the spectrum is Hock Tan, CEO of Broadcom. People outside of semis, and plenty in the industry, do not fully grasp how Broadcom has rewritten the playbook for the industry. This playbook looks a lot like private equity, befitting a mature industry – radically reduce expenses, focus on commercializing gradual improvements in product, shed businesses that are not in dominant #1 or #2 positions among a small group of competitors. Broadcom has now reached the point where there are essentially no semis companies left for it to buy, at least none large enough to move the needle on their finances, so they have taken their playbook to the enterprise software market.
That leaves room for smaller companies to replicate the model, gradually rolling up other companies in the space, and focusing on financial rather than technological improvements.
For larger companies, yesterday’s news about Intel provides a good case study – Intel is selling its SSD Nand Flash Memory business for $9 billion. Dylan Martin at CRN provides a good overview of the deal. Intel is not exiting memory, but it is shedding one of its non-core, poorer-performing businesses. The deal itself is fairly complicated, spread over multiple years, which is a smart hedge in a world in which semiconductor deals attract a lot of regulatory attention. And the implications for Intel are also fairly complicated, there are going to be a lot of people who feel that Intel is abandoning a semi-strategic asset for the purpose of ‘financial engineering’. We are not memory experts, but we tend to agree that divesting assets may not be the optimal strategy for Intel. That being said, we agree pretty strongly with the underlying principle that chip companies need to reassess their approach to products and growth.
All the chip companies, at least in the US and Europe, need to think through their strategy in coming years. As we have said, repeatedly, the number of standalone chip companies will continue to shrink. Standing still is easy now, with all-time high share prices and healthy demand and pricing. But someday that will change. Companies need to start thinking about their retirement plans – save up, bulk up or blow it all on a new bike and a leather riding suit.