X

Texas Instruments paves a solid chip path

TI's stock has already appreciated more than 60 percent YTD, but I believe that TI can continue to outperform.

5 min read
Texas Instruments is emerging as the category killer communications semiconductor supplier.

I view TI as a leadership technology company in the ranks with Cisco, Intel and Microsoft. TI continues to be one of my favorite stocks for long-term growth, and I feel even more confident about TI today than when I raised my recommendation on the stock to "strong buy" in November, 1998.

DSP (digital signal processor) technology is core to TI's strength now and in the future. DSP technology can be leveraged across TI's wireless, DSL modems, cable modems, VoIP (voice over IP, or voice delivered using the Internet Protocol) and Internet appliance businesses. TI envisions the day when every phone call--wired or wireless--all digital music and all digital cameras are based on DSPs.

A DSP is optimized to perform multiply-accumulate operations at high-data rates for real-time applications and real-world signals. A DSP performs mathematical manipulations on the incoming signal to enable the signal to be transmitted easily at low power while retaining signal integrity. The world we experience as humans is one of analog signals, like sounds and colors. Yet digital systems offer many performance advantages, including lower-power requirements and crisper, clearer signals. Because digital signals can be tightly compressed, digital systems also have more available system bandwidth.

Variable access to information is also possible with digital systems. Music in digital format, such as on a CD, can be accessed variably; with an analog tape only sequential access is possible.

For all of these advantages, the electronics world is going digital. TI is poised to benefit from this transition as the company with the No. 1 market share in DSP.

The company remains dominant in wireless. TI DSPs are used in the base stations at eight of the top 10 OEMs (original equipment manufacturers) and in over 60 percent of wireless handset basebands. The baseband is the digital portion of the wireless handset, and the base station is the wireless system transceiver that sits on top of buildings, mountains and so on.

Of the base station OEMs that do not use TI, one uses an ASIC solution and the other uses a proprietary DSP core.

TI continues to see wireless growth accelerating. With worldwide digital penetration at about 9 percent in 2000 with about 435 million units, I believe that the unit proliferation will head toward 18 percent penetration and about 864 million units by about 2003. History has shown that there has been upside to prior wireless forecasts.

Another issue in the wireless arena is a transition to 2.5G (2.5 generation) and 3G (third generation) for wireless data traffic. While many other companies are vying to challenge TI during this transition, I believe that 2.5G and 3G wireless systems are TI's to lose given that the company already has dominant relationships with the largest vendors in 2G (second generation).

In addition to wireless, TI has a broadband strategy that includes DSL, cable and VoIP.

I think that TI is unique in comparison with its competitors because it offers products that cover the full breadth of DSL offerings, including both internal and external ADSL modems on the client side and in the central office, VDSL on both the client side and in the central office, as well as related high-voltage and analog products.

TI is engaged with four of the six leading consumer PC OEMs for a USB DSL modem. The company has a new emphasis on providing full hardware and software solutions for DSL, and I believe that this will help to drive faster DSL adoption. I also believe the constraints on launching DSL will lift as operators will no longer need to visit the customer premise to install DSL.

The number of DSL lines worldwide should explode from 1 million lines today to 30 million lines in 2003, according to TI. DSL modems are required on both ends of the line, so the number of DSL modems are projected to be 2 million in 2000 and 60 million in 2003.

TI also sees cable becoming a DSP application, especially with VoIP. The company has been building its broadband capabilities through acquisitions of Amati (DSL), Libit (cable modem) and Telogy (VoIP).

I believe that having Cisco as one of its major customers validates TI's strategy. Cisco bases its IP telephone strategy on TI DSPs.

TI is continuing to focus on the attach rate to DSP sales. This metric is the rate at which TI can cross-sell analog products to its DSP customers. TI's strategy is to leverage the company's dominant market share in DSP by bundling analog parts and RF (radio frequency) parts to its DSP sales. TI has grown the attach rate from about 8 percent in 1998 to more than 25 percent in 1999. The company is tying sales person incentives to performance on this attach-rate metric and leveraging its success in DSP to sell more high-margin analog products.

The installed software base is critical to TI's continuing dominance, I think. 50,000 programmers worldwide are writing code on TI DSPs, according to TI. DSPs are typically not programmed with high-level languages (like C or Java) but instead with assembly language, where there is a smaller base of programmers. Given the enormous base of software that exists for TI DSPs, I believe that there are significant barriers to switching to another vendor's silicon.

In my view, control of its manufacturing capacity is a real strength to TI, especially as foundry capacity is tightening across the semiconductor industry. Large internal capacity gives TI an advantage with certain large customers, including Sun and Nokia. These companies feel more comfortable with TI because it controls its manufacturing capacity, rather than outsourcing this to foundries. This relationship seems to be even more valuable as the cyclical upturn for semiconductors continues.

Emerging applications--like Internet audio and digital cameras--will be future contributors to growth, and TI is positioning itself to capture these opportunities. The company has been using its TI Ventures venture capital arm to learn about emerging consumer and mobile markets. Programmable DSPs offer a compelling time-to-market proposition in markets where standards change rapidly and product cycles are short. As the digital consumer market moves toward convergence products, I believe that a standard DSP platform will be an advantage. TI is using the same low-power DSPs in digital audio players as it uses in wireless handsets, for example. Therefore, minimal incremental investment is required for TI to expand into new markets.

Since I upgraded TI's stock to a "strong buy" in November, 1998, TI has appreciated 444 percent (vs. 168 percent for NASDAQ and 13 percent for the Dow Jones Industrial Average). Though the stock has already appreciated more than 60 percent YTD, I believe that TI can continue to outperform as margins increase, broadband applications ramp, the company leverages its internal capacity in tight capacity environment, and new opportunities for DSP and mixed signal continue to emerge.

TI's fundamental business is in superb shape. I believe that this company will continue to take market share and expand into new applications.

Robertson Stephens maintains a market in the shares of Cisco Systems and Microsoft.