The Fab Owners Association (FOA) convened this month (February) for its annual review of the integrated device manufacturing (IDM) model. It’s not a meeting of hyper-happy executives. You need to go to a software conference for that kind of fun. Rather, it’s a no-nonsense review of how the IDM model is faring in the fiercely competitive business of making chips, selling them, and finding lucrative markets. Problems persist. It’s a tough industry. It doesn’t have the financial allure of fabless companies or those partying software folks. But for anyone who’s been on the IDM roller-coaster for a while, you know we still enjoy the ride.
Surprisingly enough, my overall take-away from the FOA meeting is that things are looking pretty good. Of course, I’m a PR person and we’re supposed to be positive or we need to find another career. But my opinion is not based on my optimism, but that of the participants.
Some quick takeaways from the agenda:
7% Outlook for 2014
Bill McClean of IC Insights sees the 2014 worldwide semiconductor market garnering $350.7 Billion, for 7% growth over 2013 revenues. His firm pegs USA GDP at a 2.7% growth this year, although his caveat is that the IC industry as an ‘economy driven industry model” whereby folks buy electronics based on how they feel about the economy. (I disagree.) While 2.7% may seem low, he looks at the European Union to come in at 1.1% and Japan at 1.3% GDP. Turning from the macro-picture, Bill noted the strong promise of devices relying on analog and power technologies.
Analog and power management. Remember both those technologies. They play a role in every discussion about ongoing and future opportunities.
China, Innovative Friend or Foe?
Handel Jones, PhD, and president of International Business Strategies, gave an overview of the China semiconductor market. Aside from praising the Chinese IC industry’s strides, he had some interesting insights into the IDM model. He believes that that control of wafer fabs has advantage for analog, power and what he termed ‘specialty areas’. And he commented that the fabless model was over-rated for many markets. Buy his book ChinAmerica and you might learn more.
The high-growth markets he named were MEMS sensor, RF and analog-based technologies.
Dr. Jones also noted that semiconductor innovation has gone global and that technologies based on GaN show strong promise.
Innovation – Alive and Thriving or Slowing Down?
Innovation is a tough topic in the industry. There are some who believe that innovation has slowed down. Nothing could be further from the truth. Five years ago, analysts were projecting market growth for digital/video cameras. Now that market’s been absorbed into phone market statistics. That’s phenomenal innovation in a very short timeframe.
I maintain that some of the best technology innovation has come from the packaging/assembly, equipment and materials industries. Consider the intelligence it took to manufacture machines that churn out 20nm chips. Or 3-D packages. Or semiconductor chips you can swallow for an in-body look at your health. Or the materials used to deliver those chips. Innovations like these occur all the time. They just don’t garner the consumer media’s attention compared to the latest smart-phone software app. And those apps, by the way, come to life on a slice of silicon.
Your Chip-Based Home
Joanne Itow of Semico sliced the market a bit differently from earlier speakers by looking at the number of electronic devices in a household. Semico counts >70 of them. The affordability of consumer electronics presents an interesting dichotomy created by technology. Consumers want more features in the electronic devices in their home, but they don’t want to pay a high price to get them. Nothing new there. However, that desire presents an opportunity for mature technologies and IDMs. IDM manufacturing shaves costs by having high-volume manufacturing and design under one roof. And, price has been a key factor in the growth of MEMS, 3D and other technologies that don’t require ultra-deep submicron technologies.
Joanne noted that while 200 mm wafer fabs still produce the majority of chips, 300 mm fabs produce more revenue and demand.
Conflict Minerals. No, Not Diamonds.
Ron Jones on N-Able Group, outlined requirements for reporting conflict minerals used in the semiconductor industry. “Conflict minerals” are defined as those coming from places like the Democratic Republic of Congo. Tin, tungsten, tantalum, gold are just four metals from conflict minerals. Semiconductor “DNA” will need to be identified and reported in SEC filings to ensure that minerals are being mined in a responsible way. Sounds tough to me.
Fab Manager: Know Thy Markets
No, fab managers do not have to become marketers, but they need to know the markets suited to their manufacturing. I’m a big believer in the IDM model because manufacturing is at the heart of it. By owning design and manufacturing, companies have the tremendous benefit of product and market differentiation; design-to-delivery timeframes; brand appeal – which includes “apps” (or, applications); and that, in turn, results in sales and revenue.
The problem is that manufacturing has tarnished the IDM model. There are many reasons. One factor is that CEOs do not have the background to run manufacturing…or the discipline to control it. Yet, many companies have been successful as IDMs. Linear Technology is one. Samsung is another. There are plenty more.
The IDM model, which is what we call an ‘aggregated’ model, ties in with the consumer desire for more apps at a lower price. By designing chips in-house, like IDMs do, companies can deliver what consumers want. This is exactly why we see Google and other companies outside the semiconductor industry taking an interest in chip design. While they may not become IC manufacturers, IDMs have an opportunity to refine their business model to capture new business, or lose it.
Thanks to ARM and other IP suppliers, even “small” companies can design many of their own chips and turn them over to a manufacturer for fabrication. I see IDMs as being able to fulfill this demand because there are more mature technologies than new ones.
Collaboration Key to IDM Success
Kudos to Applied Materials/Infineon; MEI/Anadigics; Semplastics/Intersil; Applied Materials/Polar Semiconductor; MAX/Sunpower; Busch Vacuum/Freescale; and Eyelit/MAX. These companies were the highlight of the FOA conference. They presented a probem, defined it, and showed how they solved it, along with the data to prove it. At other conferences, vendors will talk about a client or two, but the client is typically not in the audience. This model not only showed attendees how collaboration solves problems, it demonstrated how collaboration results in efficiencies that translate into dollars saved.
The IDM model has a lot of problems, like every other business. And, like every other business model, there are solutions waiting to be found. If you’re in the IDM business, you know it’s a roller-coaster ride. So what? The roller coaster is the best ride in the park.
~Barbara Kalkis, Maestro Marketing & PR(sm)