Texas Instruments has exceeded Wall Street's expectations for the past year, thanks partly to the increasingly diverse array of devices that use its chips. As sales of new MP3 players, digital TVs, and advanced cell phones have mushroomed, TI (TXN) has been able to offer an irresistible combination of cutting-edge chip design at a reasonable cost -- a reflection of its unwavering commitment to research and development.
Unlike many chipmakers, TI maintained its level of R&D investments throughout the economic downturn. It then hiked R&D spending nearly 10% in 2003, to about $1.75 billion. And TI will increase that budget again this year, says Rich Templeton, executive vice-president and chief operating officer. (On Jan. 15, TI announced that Templeton will replace Tom Engibous as CEO and president effective May 1. Engibous will continue as chairman.)
Why the extra emphasis on R&D? The consumer-electronics devices that are powering the current upturn in semiconductor demand require smaller, less power-hungry, and more complex chips than ever before. Although TI reached its longtime goal of shrinking some chip combinations for devices such as cell phones down to a single chip in 2003, customers such as Nokia (NOK) are far from satisfied. They want more functions on a chip, at a much lower price.
TI has made steady progress toward that goal, though as chips and production methods have grown in complexity, the task has become much harder. What kind of R&D commitment will it take to keeping designing better chips that cost less? And which areas of innovation will be hottest in the near future? Templeton discussed these and other issues with BusinessWeek Online reporter Olga Kharif on Jan. 9. Here are edited excerpts of their conversation:
Q: What's different about this semiconductor upturn?
A: We're in a post-PC era. We're in a world that's going to be driven by broadband applications at home, by new wireless applications, and by digital consumer products. We are embedding semiconductors deeper into industrial, medical, and automotive devices. And that broad set of industries and uses will be the driver for this semiconductor upturn -- quite possibly for the next 10 years.
Q: What are the ramifications for chip suppliers?
A: That's the beauty of this post-PC era: It's demanding a broader set of chips -- in particular, more DSP [digital signal processors] and analog chips, which are used in cell phones. But name any growing market -- broadband, wireless, digital consumer devices -- and fascinating things are going on.
Q: Which are the hottest areas of innovation right now?
A: There's a handful of important new products. In cell phones, for example, we're going to see chips based on new network standards and offering new data-receiving and -sending capabilities. Lots of innovation is also happening in broadband and in voice over IP [VoIP], a technology for making phone calls over the Internet. Digital TV is one of the hottest markets right now -- and more innovation is happening there. MP3 players, digital cameras, digital radios -- they're all pretty exciting.
Q: In this post-PC era, have the rules of the game changed for the chipmakers?
A: The list of companies that have skills in DSP and analog components, as TI does, plus the ability to integrate different technologies -- wireless, broadband and others -- has gotten considerably shorter. It's just a more complex business. To supply the next-generation broadband market, you have to have chips for the DSL [digital subscriber line] modem, for the wireless LAN [local area network] -- and the intellectual property to put that all onto a single chip. Not too many companies can do that.
The same holds true with chips for next-generation cell phones, which need to combine multiple chips, including wireless technology for Bluetooth, wireless LAN, or global positioning systems.
Q: What do you expect to happen with the R&D budgets of players that stay in the game?
A: My experience is that constant innovation is rewarded and that lack of innovation is usually penalized, so I see R&D dollars continuing to grow, along with revenue. At TI, our R&D will be larger a year from today than it is now. Mind you, our sales continue to grow at a double-digit rate year over year. [Eventually], I would like us to get to a point where R&D is back down to about 15% of revenue [vs. about 19% last year].
Q: For many years, TI has talked about combining a number of different chips onto a single chip as being the holy grail. You achieved that in 2003. What's the next challenge?
A: Yes, we introduced a single chip that includes Bluetooth and other functions in the middle of last year. But the capabilities of cell phones, for instance, continue to grow at a rapid pace. So every time we integrate more functions on those devices, the number of things that our customers want to put on a chip continues to grow as well. This is a never-ending process. Customers always want more functionality for less power and less cost.
Q: How do you reduce the cost of chips?
A: I don't see anything fundamentally changing on that score. We are increasing the diameter of the wafers from which chips are cut. And we expect to move to smaller chips over the next 10 years.
Q: Analysts say that since consumer-electronics makers are facing more competition and shrinking margins, chipmakers will need to help them lower costs and differentiate their products. How can you help?
A: What [our customers] want is leading-edge capability. They want to know that their supplier is bringing them new capabilities first. So you have to be first to market. And then they want some room left for them to do differentiation -- for instance, by adding extra features.
Q: In terms of the features, what do the end users want?
A: Consumers want products that are easy to use and that do the job well. I know it's a simple statement. But one of the weaknesses of the technology industry is that it tends to sell based on technology or features, when consumers just want to turn the device on by hitting a button.
I think that companies that do the best job of making their products easy to use and intuitive will win. In the case of broadband and wireless LAN, make installation simple. Companies that do that will be rewarded.
Q: How can chipmakers help with that?
A: Look at a cell phone 10 years ago. Back then, the phone had to be plugged in. Its battery life wasn't very long. It was bulky. What did we do? We reduced the number of chips needed to perform all the functions within a phone, which reduced the power consumption. We also put much higher-performance chips inside, which allowed cell-phone companies to make the phones easier to use. That's why cell phones have become as big a market as they are today.
Looking ahead, we'll see the same thing happen with broadband and wireless LAN. These are wonderful technologies, but it's not that easy to get them installed. As we make chips more powerful and less power-hungry, equipment makers will be able to use the software and the functionality to make installation routines easier. They'll perhaps have more automatic setups, so a lot of things that have to be done with an instruction manual today become intuitive -- and that should fuel adoption of these technologies.