Technology

Taking the Pulse at Texas Instruments


CEO Templeton discusses the company's desire to capture even more of the analog chip market and hints at future involvement in advanced medical devices

The future's looking pretty good at chipmaker Texas Instruments. As the main chip supplier to the makers of cell phones, it has enjoyed steadily growing sales in recent years, from $12.6 billion in 2004 to $14.3 billion last year, while profits have zipped along from $1.8 billion to more than $4 billion over the same period.

While its supply contracts with mobile-phone manufacturers including Nokia (NOK) and Motorola (MOT) grab all the headlines, the bulk of Texas Instruments' business comes from chips that are nowhere near as attention-getting. Accounting for about 40% of the company's sales in 2006 are its so-called analog chips. Analog chips measure "real world" inputs—from the temperature in a room or the sound of a voice to the touch of a button—so they can be converted into a digital signal that a computer or other device can understand. Texas Instruments (TXN) is easily the world's biggest supplier of these analog chips.

They may not sound as advanced as the newest Intel (INTC) microprocessor, and in many cases they aren't built on the latest semiconductor technology, but as long as human beings use gadgets of practically any stripe—cell phones, camcorders, PCs—there will be a need for analog chips like those that Texas Instruments makes.

On May 14, Texas Instruments Chief Executive Richard Templeton sat down with BusinessWeek Senior Editor Dan Beucke and Senior Technology Writer Arik Hesseldahl to talk about the state of the company's many businesses, from the analog operation, to its Digital Signal Processor (DSP) chips that are widely used in wireless phones, to opportunities with medical devices.

How are the profit margins on analog chips?

The profit margins, depending on which part of the analog market you're in, are very good. And they [analog chips] use older [manufacturing] technology, so they generate very good cash flow. There is some leading-edge technology as well, but they tend to be a great business model. So analog tends to be a great business, but overlooked.

Intel has been talking lately about introducing either a new chip design or a new manufacturing technology every year. It sounds as though with analog it's a different story.

To put it in perspective, we've got some analog parts that are 25 years old. We've got some DSP parts that are more than 20 years old. But the newest ones, say the DSPs that go into the new Voice-over-Internet-Protocol (VoIP) base stations and gateways, they're ramping at, the latest, 65 nanometers, and a family that will follow at 45 nanometers. So we'll bring the latest and greatest out, but what's different is the older ones tend to stick around for a very long time. In the wireless phone business we'll have some products that turn over almost as fast as Intel's microprocessors.

One company that has buzz right now is Apple, and its forthcoming iPhone device. Since you're so active in the wireless handset business, what's your impression of the device and how it may shake things up?

First, I think it's great when we have customers in a market where people invest in research and development, who innovate and grow. And I put that up there in contrast with the PC industry. PC companies invest what amount of their revenue in R&D?

The biggest ones, not very much.

The fact is, if it's a point or a point and a half, that's very little. And so when people sit back and ask why there isn't anything new, it ought to be an obvious answer. The PC companies aren't investing. When I say I think it's wonderful that there are companies like Apple (AAPL) investing in R&D, improving the user interface, new features, [and] companies like Nokia, Ericsson, Motorola all working on R&D, that's all good stuff because when those people are doing that, it's going to lead to new ideas and growth. And the last time I checked, chip companies can't grow if the people they sell to are not growing accordingly.

Will you supply any components to the iPhone?

Apple is a good customer of ours, and they also like us to be very private about who supplies what.

Who are Texas Instruments' biggest customers?

Nokia is still the largest. We only disclose the ones that amount to larger than 10%, but there are companies like Motorola that have a chance of becoming very large as well, so we end up with a pretty broad range of large customers.

Much of the buzz right now is around VoIP. Having already disrupted the traditional telecom business, it's now poised to start affecting the wireless business, which goes squarely to what your customers have to think about. What affect is VoIP having on your business?

TI is probably the largest supplier to VoIP, whether it's to phones or things that no one ever sees, like the gateways. A lot of the operators want to take their voice traffic off the old TDM lines so they will have gateways as close to the edge of the network as they can get. We love that because conversion of the old lines to VoIP is all about signal processing, which means farms of DSP chips. Keep an eye on the cable companies switching large areas of neighborhoods and offering voice along with TV and Internet. We've got a lot of the voice processing that goes into the cable modems. So there's a lot of interesting dimensions to VoIP, many of which the consumer never sees. That's going to continue. And stay tuned for VoIP in the wireless domain. What people are talking about on 4G [fourth-generation] networks is treating the network as a bit pipe and turning voice into just another application on the phone.

What business should Texas Instruments be in that it's not in yet?

If there is one thing we're saying right now is that analog has become an even more important opportunity than before. There's a lot of growth potential that we can still find. And it doesn't make the front page of an annual report, and it's not this big single killer application. It's just a very large, very diversified, very good opportunity. But we're beginning to see a lot of emerging and new opportunities in the same way that semiconductors have revolutionized computing and communications. Think about the medical industry. It's about 15% of the GDP in the U.S., but the impact of semiconductors in medicine is near zero.

In what way?

Take diagnostics and laboratory equipment or ultrasounds. We have customers who have developed ultrasounds in a 10- by 12-inch portable package so you can get it closer to the patient. Then there's high-end medical imaging. Doctors want almost HDTV-quality images because they want to do image-assisted surgery. Think about the opportunities for DSP and analog parts inside that equipment. Then there's handheld devices like glucometers that help check and measure daily insulin levels for people with diabetes. The opportunity for that is going to do nothing but grow.

Who are you working with on these ideas?

I'd guess we have 8 to 10 people from different product areas that are gathered up into a medical team. Plus, I have a smaller group that I get together with a few times a year to try to understand where we're going. We've had those groups at a half-dozen medical schools and bioengineering schools. We want to get a snapshot about the tough and gnarly problems they're looking at 5 and 10 years down the road. We're staying closely connected with a lot of the smaller customers, very innovative companies, a lot of them not public. Few even make it public because they get acquired before they get to that point.

Do you see yourself making acquisitions in this area? Or is it too soon to say?

It's too soon to say. You've watched us in the DSP and analog world make smaller acquisitions because we really respect very targeted teams that have deep and specific knowledge, so we're not afraid to do something along those lines.


Later, Baby
LIMITED-TIME OFFER SUBSCRIBE NOW

(enter your email)
(enter up to 5 email addresses, separated by commas)

Max 250 characters

Sponsored Links

Buy a link now!

 
blog comments powered by Disqus