|
|
 |
|  | ANALOG DEVICES AMPLIFIES ITS SIGNAL | | | With a new partner strategy in place, ADI wants to rewrite the rankings in the DSP market. | | | By Richard A. Quinnell | |
|
Digital signal processing is a key technology for applications ranging from cell phones to multimedia to appliance motor control. Texas Instruments is the 800-pound gorilla that dominates the market, yet the much smaller Analog Devices has positioned itself to challenge that dominance. A key element of Analog Devices' strategy has been to form design chains that let the company capture new markets just as they begin to emerge.
 Belying its name, Analog Devices, Inc. (ADI), based in Norwood, Mass., has had a presence in the digital signal processing (DSP) market for more than 15 years, building on its analog signal processing foundation. Although more than 80% of its revenues still come from its analog products, according to company financial reports, its activity in the digital world is a natural extension of the company's core expertise.
ADI is positioning itself to be at the forefront of signal processing technology, both digital and analog, in anticipation of a significant industry shift. As company president and CEO Jerald Fishman tells investors, "We believe that signal processing is becoming the primary growth driver for the semiconductor industry. There is little doubt that high-performance analog and DSP will be the enabling technologies at the heart of high-growth opportunities over the next five to ten years."
ADI's goal is to seize the lion's share of these opportunities just as they start to emerge. The company's analog reputation is well established, but to control these future markets, it needs a compelling DSP strategy, as well. It has one. ADI is using a two-pronged approach that lets its efforts in the horizontal DSP market help it identify emerging markets. The company will then create design chains that allow it to develop and support more-vertically-focused DSPs.
 ADI is not yet in a position of dominance in DSPs. Texas Instruments holds that position. In 2001, merchant sales for digital signal processors reached $4.26 billion, according to Tempe, Ariz.-based market research firm Forward Concepts. By the firm's calculations, Texas Instruments (TI) leads with 43.5% of that market, whereas the challenger, ADI, took home 8.2%, placing it fourth, after Motorola and Agere (see the chart, "Programmable DSP Chip Market Share"). These numbers don't seem out of the ordinary: The ratios reflect the relative size of the two companies. TI reported 2001 revenues of nearly $8.4 billion, compared to ADI's $1.7 billion.
But ADI is in a position that is stronger than it first appears. The DSP sales numbers are skewed by the size of the cell phone market, observes Will Strauss, president of Forward Concepts. "TI has 50% of the market for base-band chips, including DSPs, for cell phones," says Strauss. "If you strip out the cellular portion, TI had about $600 million in the non-wireless DSP market whereas ADI was in the $350 million range." Remove all the other vertically focused DSP applications that both companies target, such as brushless motor control, and the two companies become comparable. "The general-purpose DSP market is really a two-horse race between ADI and TI," says ADI's Gerald McGuire, general manager of ADI's media platform and services group, "with each having 30% of this $500 million market."
 A Twofold Strategy
Although being a top player in the general-purpose DSP market is advantageous, ADI's plans don't stop there. It wants what TI has: huge target markets. The first step is getting into a new market on the ground floor. "Our horizontal strategy provides early visibility into emerging markets and a chance to work with the key players," says McGuire. Strauss feels that the strategy has a good chance of providing the needed insights, noting that "ADI is still considered the premier A-to-D company in the world. That gets it in a lot of doors."
The trick, of course, is to convert that early entry into market dominance as an emerging market matures. Stephan Steyerl, an ADI strategic marketing manager responsible for one such conversion, explains: "Lots of companies wait for a market to fully emerge and then start gaining expertise. We may not have a product optimized for a market to start with, but we begin working with customers, IP providers and other parties to develop that expertise. Our organization gives us the freedom to explore opportunities without short-term revenue pressure."
The method for tackling an emerging market varies with each opportunity. "There is no easy answer," says McGuire. "Usually the market characteristics dictate the right approach. In some segments, tier one customers act as early adopters and invest in product development. Then, second-tier customers follow their lead, using the resulting standard products." In other segments, ADI optimizes its design chain by working with other providers. "We identify an IP vendor and start developing a close, long-term relationship," says Steyerl.
 That's the approach ADI took with the refinement of its TigerSharc DSP product for use in third-generation (3G) cellular telephony base stations. "We worked with our existing 2G customers to define what would be the most efficient product," says Steyerl, "and then partnered with IP developer Delphi Communication Systems to adapt the TigerSharc to meet those needs."
"ADI asked us to implement a 3G base station reference design," says Doug Schute, president and CEO of Maynard, Mass.-based Delphi. "We created a two-board set and then optimized our software for the TigerSharc." But this was not a simple design assignment; it was a partnership. "ADI worked with us as peers, rather than have us simply lob a reference design over the fence." The two companies worked side by side, developing the reference design and concurrently optimizing the software as well as the TigerSharc for the target application.
Managing the Endgame
|
|
|
Nothing lasts forever. As any marketing manager will quickly agree, promising products fizzle, and cash cows eventually run dry. With a product developed entirely in-house, backing out of a declining market is reasonably simple. When products are the result of a closely linked design chain, however, it gets a little more complicated. You have to consider the impact on your design chain partners or risk losing them on future projects.
"You can spend a lot of money chasing these solutions," says Doug Schute, president and CEO of Delphi Communication Systems, ADI's partner in the third-generation base station market. "If the DSP guys change direction, you can get left out in the cold. That goes all the way up and down the design chain." As a result, partners in a design chain can risk close cooperation only when there is mutual trust. "You want to be comfortable with your partners," says Schute, "if you're going to hand the family jewels to them."
Carefully framed contracts can help prevent the actions of one partner from damaging the other, but only so far. "Agreements address each partner's rights," says Schute, "but in the end, they can't address issues such as trust and ethics. There has been a lot of exploration of partnerships in the last five years, and a lot have failed miserably. The ones that work are based on mutual respect, trust and honesty."
Recognizing its responsibility to its partners, ADI does not expect to make a fast exit from markets it enters. ADI's approach gives its design chain partners a chance to react and form their own exit strategy, which is all that companies such as Delphi ask. Says Schute, "You have to be able to absorb changes and move on."—R.A.Q.
|
|
|
|
|
Design Chain Serves Dual Need
The formation of this design chain had two purposes. One was to jointly develop an optimum system design for ADI's 3G component customers. "The working relationship had to be extremely close," says Schute. "It was as if our team were an extension of their organization." As a result of that closeness, Delphi was able to affect ADI's own efforts. "The fundamental architecture of the TigerSharc had been defined early on," says Schute, "but we could help refine it. We were able to add some special instructions, such as Viterbi decoding [an error-correction capability for phone transmission], to make it more efficient on the base station algorithms."
The collaboration was clearly successful. "The TigerSharc is considered the only chip around," says Strauss, "that does all the essential signal processing for base stations in a single chip. TI will be announcing its own product, but it's a three-chip set."
Product refinement was not the only goal, however. The design chain ADI forged with Delphi extends beyond the product's creation into customer support. "For the 3G base station application," says Steyerl, "we have engaged with several tier 1 customers ranked among the 10 or so companies that we think will succeed in capturing 90% of the 3G base station market. There may be 20 to 35 tier 2 customers, as well. ADI isn't structured to support its tier 2 customers as fully as it supports tier 1, but our partnership with Delphi helps us ensure that those customers will get the support they need."
Delphi serves as a consultant to these additional ADI customers, an arrangement Delphi welcomes. "Generally speaking, the thing that constrains you as a DSP supplier," says Schute, "is the amount of support you can provide, how many people you can apply. And there are never enough. So ADI has segmented how it supports its customers. It has kept tier 1 in-house and directs the second and third tiers to Delphi." Schute fully supports the segmentation. "In some cases, we have been brought in to talk with the tier 1 customer," he says. "But many large customers want a single point of contact."
Working Closely Carries Risk
This type of segmentation is possible only because of the close working relationship the companies had when developing the reference design and refining the TigerSharc. "To provide this kind of support," says Schute, "Delphi has to be as knowledgeable as ADI on the DSP, the tools and so on."
Over time ADI will broaden its own customer support for vertical products, using a two-part scheme. "The first step is to fine-tune the product to address tier 2 needs," says ADI's Steyerl, "and the second step is to add to the support infrastructure to handle tier 2 customers." The actual effort varies with the market and product complexity, Steyerl adds. "It depends on whether the product is a breakthrough technology or a derivative."
 This dichotomy in customer support is an artifact of ADI's blended market strategy. The extra effort is worth it, however, according to McGuire. "It's unwise from a financial standpoint to become dependent on one or two narrow market segments," McGuire says. "Vertical markets can see—and have seen—high volatility. The strategy of ADI is to maintain a healthy mix of vertical and horizontal market businesses."
With this one-two punch of a horizontal business leading into vertical markets, ADI has positioned itself for early access to emerging markets. With the formation of design chains to provide expertise and customer support assistance, ADI has put itself on an even footing with TI in the competition for those markets. And if it manages to capture the right one, ADI could become the next 800-pound gorilla in the DSP market.
|
|
ABOUT THE AUTHOR |
|
 |
|
Richard A. Quinnell (rquinnell@designchain.comm) has been a technology journalist for more than 14 years with such publications as EDN, Electronic Business, and Computer Design. |
|
Back to the top
|
|
