Microsoft, CMGI and Cisco Systems are just a few of the industry players
we've interviewed. Check out the StockTalk archives.
Does eBay have the best online community? Its VP of Marketing thinks so.
CDnow's CEO tunes in to the company's music destination.
Is Abercrombie & Fitch a high-quality American classic? Get its Investor
Relations Director's perspective.
TMF Interview With Texas Instruments Chairman, President & CEO Tom Engibous
February 3, 1999
With Yi-Hsin Chang (TMF Puck)
Dallas-based semiconductor manufacturer Texas Instruments (NYSE: TXN) is the leader in digital signal processors (DSPs) with a 45% market share. DSPs, which convert analog signals such as light and sound into digital signals, are found in everything from modems, VCRs, camcorders, cars and cellular phones. TI has completely remade itself by divesting 13 businesses and acquiring 10 others in the last two years. It now focuses primarily on digital signal processing and analog technologies.
TMF: One of the first lines I read on your website said, "TI is a new company." Do you think the company is misunderstood, and if so, what are the most common misconceptions?
Engibous: I think that's separated between what is the internal knowledge and the external perception. I think internally all of the TIers know exactly what our objectives are and the direction, and there's not any misconception at all. In fact, one of the benefits of being far more focused on programmable DSP and analog is that we don't have to deal with what's most important between radars and missiles and notebook computers and so forth. It's a very, very clear direction.
I think outside, the external perception in the investment community, has been very understanding, very up-to-date, and obviously what's happened to our stock price is a direct result of that. If you go to the general public, TI is never a highly visible company because a very small percentage -- probably about 5% to 10% -- of our revenues even reach the general public, and that's mainly through our calculator line. So it's true in the general public that the perception of TI being a new or different company probably lags by quite a bit.
"DSPs go into literally thousands of applications. That makes it very difficult to mount some type of 'TI Inside' type of campaign."
TMF: Intel's (Nasdaq: INTC) been doing a lot of marketing of late. Do you see yourselves doing anything like that?
Engibous: Well, I think that Intel has an enviable position of having a near majority, not quite, but a near majority, in the PC space. One of the advantages -- the biggest advantage maybe, and also in this instance a disadvantage -- is that DSPs go into literally thousands of applications. That makes it very difficult to mount some type of "TI Inside" type of campaign, so I don't think that is something we're going to be seeing in the near term anyway.
TMF: Three years ago, DSP and analog components made up about 20% of your semiconductor revenue. Today they make up about 58%. The company has said that it expects that figure will grow further. Exactly how much further, and what's the ideal breakdown percentage-wise in your view?
Engibous: Well, we don't have an ideal number, and it'll never be 100% because there are some businesses that are parts of TI today that are great businesses. I mentioned a non-semiconductor example -- calculators, which is just a great business for us so it won't be DSP. In semiconductor there are some supporting businesses, like our ASIC [applied specific integrated circuits] business, as an example, where a great deal of the DSP business that we ship today is actually designed using our ASIC business skills. So we want to continue to compete in that business outright in order to ensure that we have the greatest ASIC capability to use with our DSP. But with all that said, it's clear simply because of the fundamental growth rates of DSP and analog that that percentage will grow far more than 60% over the next few years.
TMF: What is your strategy going forward? Will you continue to be aggressive in making acquisitions, and are there still businesses you'd like to divest?
Engibous: I think from the divestiture standpoint, the businesses that we now have are all leaders in their various spaces, and they meet our criteria for value, stability, and growth. I'll never say that we won't ever make any more divestitures, but certainly of all the big ticket items that we've done, we see that as coming to a close. The other side of that coin, though, I'll say with an exclamation point. We will definitely continue to be aggressive in strengthening our position in programmable DSP and analog. Acquisitions, as we've viewed them in the past, will be a tool for strengthening those strategies, and I clearly believe we will continue to do that.
TMF: You've done a high number of acquisitions in the last two years. Do you think that will be normal for you going forward?
Engibous: I don't know that we even look at it that way. I would say it this way: We certainly don't need to use the acquisition tool as a means of revenue growth. The DSP market is growing greater than 30% per year. In fact, in the semiconductor industry, last year was a negative growth year again, and our DSP business grew at 29%. The revenue growth potential is there with our existing portfolio. Where you see some of the small acquisitions being made are in pieces of the strategy that give us a much stronger position, predominately in the software space, both in development tools for programming DSPs and application software to be used on our DSPs.
TMF: Competition in the DSP market is fierce, especially with Lucent Technologies (NYSE: LU) teaming up with Motorola (NYSE: MOT) and Analog Devices (NYSE: ADI) reportedly linking up with Intel. What will give TI an edge over its competitors, and how will you retain and, in fact, gain market share?
"The DSP market is growing greater than 30% per year. In fact, in the semiconductor industry, last year was a negative growth year again, and our DSP business grew at 29%."
Engibous: Well, you are exactly right. Competition is extremely fierce. It's not a secret that the programmable DSP and analog space is a lucrative space both in terms of growth and value, so it attracts just about anybody that wants to drill a well somewhere. That won't change, and we have to recognize that. It's also true that a number of people have linked up trying to gather strength together through joint ventures or partnerships and so forth. Those all remain to be seen.
The history of those [joint ventures] is not very good, and you can list those even down to power PC and so forth. The other thing is, as each of these are being formed on the back of what we believe to be some degree of weakness in their own positions as we've continued to gain market share over the last number of years. But we don't take our competition lightly, so we'll continue to be extremely aggressive in introducing the highest performance products, the widest portfolio of products and back that up with the lion's share of the installed software base, and so far that's been a very competitive and successful strategy.
TMF: With so much competition, are you concerned about pricing pressures that could cut into your profit margins?
Engibous: One of the nice things about the programmable DSP and analog market is that it is less price sensitive than most of the other businesses in this industry, albeit for two different reasons. The differentiation in analog is very simply complexity -- literally thousands of parts, very difficult to duplicate, and because of that complexity, margins tend to be relatively high. Historically that remains a very different sort of business.
In programmable DSPs the differentiator is installed software base. DSP is extremely difficult to program, so once people make an investment in software, once they choose an architecture and write software, the tendency is not to change. That ability to capture the installed software base has proven to be a very, very good differentiator in keeping the pricing pressures reasonable.
Now on the other hand, it's to our benefit to continue to drive the cost per MIPS [millions of instructions per second] down because it introduces DSP into new application spaces that have been previously done by other approaches. So it's a situation far better than the rest of the semiconductor market particularly when you are in the lead of an installed software based space. But we will continue to use technology to drive the cost per MIPS, which is a computing measure, down such that we can attack far bigger and an increasing number of sockets.
"It's not a secret that the programmable DSP and analog space is a lucrative space both in terms of growth and value, so it attracts just about anybody that wants to drill a well somewhere."
TMF: NEC's recent adoption of TI's VDSL (very high bit-rate digital subscriber line) technology is a significant step forward for you. Could you talk about your plans for this market with reference to the deal?
Engibous: Well, I can talk about this market because it's an exciting market. The entire Internet access market has gone through a significant change over the last few years. It used to be what people would call a hard-wired or hard-coded market where no programmability was put into Internet access devices, which at the time had been mainly analog modems. Over the last couple of years that completely flipped, and essentially now all modems are programmable DSP driven. Obviously that's had a major turnover in who the leaders are in this industry, but in total the biggest significant change has been a move to programmable DSPs driving Internet access devices for both upgradability future proofing and improved performance over time -- by downloading new code over the Internet you can change the performance.
As we go forward, the next generation of Internet access, or at least one of the main ones, is DSL, and today TI's C6000 happens to be the only single chip DSP that can do a fully programmable ADSL solution. So we've taken that performance, we've coupled it with the world's best DSL software -- that is in our purchase of Amati Corporation -- and we've put that together, and it's getting excellent reception by the leaders providing this technology both from an infrastructure standpoint and to the client side.
TMF: What would you say are the most exciting new products in your pipeline?
Engibous: Well, I don't want to talk too much about the products in the pipeline. I can tell you that the most exciting product line out today is definitely our C6000, which now consists of a number of products. In terms of measuring its acceptance we have a way of measuring design wins by measuring the number of linker sales -- it's a small software tool that if you are going to write code you buy it and you write code. The measure of linker sales on our C6000 dwarfs -- and I do mean dwarfs -- the success rate of any other DSP we've ever introduced. The move to this new architecture, the 2 1/2 year lead in performance that this product has produced has generated a tremendous amount of excitement, and we're seeing it in early design wins that ultimately will lead to, we think, really good, strong business success.
"We don't take our competition lightly, so we'll continue to be extremely aggressive in introducing the highest performance products, the widest portfolio of products and back that up with the lion's share of the installed software base."
TMF: Now you've been CEO of Texas Instruments for almost three years. What would you say are your proudest accomplishments?
Engibous: Oh, I'm just one of thousands of TIers working on this. I think that as a team we feel very good about the focus we now have on this company about being the world's leader in programmable DSP and analog. It's been a difficult task to move from a very diversified, not only in product but industry, to this focus, and I think that all of us feel extremely good about the position that we now have.
If we view where electronics is moving, and you can't talk to anybody that doesn't say that we're moving to the communications era, and communications means more bits down the same pipe, and every time you want to send more data down any pipe whether it's wireless, cable, or telephone lines, or fiber, that is a DSP function on both ends of the wire. With just that statement alone, we find ourselves in a position that we are really excited and proud about.
TMF: Thank you for taking the time to talk to us today.
Engibous: Yi-Hsin, thank you very much.
Got an idea for StockTalk? Who would you want us to interview and what should we ask? Drop us a note at StockTalk@fool.com.
Would you work for a bunch of Fools?