The mainstream media may call it a digital age. But today’s gadgets still need integrated circuits that can transform analog signals—which convey information about “real world” phenomena like temperature, pressure, sound, and speed—into digital form.
Linear Technologies is one of the leading companies designing, manufacturing, and marketing a broad line of standard high-performance analog integrated circuits as well as devices that control power and regulate voltage in electronic systems. These products can be found in cable modems, cell phones, disk drives, radar systems, satellites, and scientific and industrial instruments. About two-thirds of Linear’s sales come from overseas customers, and its two largest segments are the industrial and communications markets.
Much of the company’s success comes from its sales force, which includes many engineers who can help customers create solutions. According to one story, a Linear salesman had met with a customer who had an idea for a camera phone. On the plane trip back, the salesman was trying to fall asleep, but he couldn’t figure out how to create a camera flash that could work with a cell phone. He closed his eyes, and as the plane passed through some clouds, the light flickered on his eyelids like a strobe. By the time he landed, he had figured out how to substitute a strobe light for a typical camera flash, and the camera phone became a reality.
Urban legend? Possibly, but you get the point.
Despite this expertise, 2007 presented some challenges. While its stock closing price improved, Linear lost some ground in its employee growth and debt to equity ratio, as well as in the number of bonus points based on Electronic Design’s annual Reader Survey. These factors contributed to Linear’s slide from its rank at 57 in last year’s list to 60 in this year’s list. 2008 offers some hope, though.
According to World Semiconductor Trade Statistics, the analog segment will grow by about 10% for 2008, versus 7% for the overall semiconductor market. In 2007, growth was flat. While 10% isn’t in the 15% to 18% growth range that the analog segment has shown in the past, it’s better than flat. Because the analog market is more diverse and fractured, the ups and downs tend to be softer than the DRAM and digital segments.
Like other companies in the industry, Linear authorized the repurchase of $3 billion (around 30%) of its outstanding shares in the middle of last year. At the time, the company had internal discussions about whether it should buy all the stock back and do a private equity buyout. However, Linear felt that buyouts were for companies that needed to be strategically fixed and then sold again.
The company also felt that it wasn’t a very good candidate for a private equity buyout, and there weren’t many parts of the company that weren’t working well. So, Linear decided to keep doing more of what it was doing. In addition, the high-performance analog space isn’t very conducive to startups backed by venture capital (VC), as the number of customers and unit sales are relatively low and the return on investment period may be too long for most venture capitalists.
In Linear’s early days, military and space products were important to sales. The focus then shifted to industrial and then to automotive and finally to communications and consumer products. High-end consumer products, mainly cell phones, drove Linear’s growth from the dot-com bubble burst to a few years ago, but now those markets are becoming more and more commoditized.
Currently, the industrial and automotive segments are back to being most important. Automotive has cycled back from 3% to 4% of sales a few years ago to 10% currently and now tops $100 million. The military and space segments, driven by a lot of satellites going through their life cycles, is starting to grow again as well.
The company also has a strong presence in the power-over-Ethernet segment. Its first generation of products was introduced three to four years ago, and second-generation products are currently being introduced. Although Texas Instruments has depressed prices in the segment as per its strategy, Linear feels it has the better product.
Linear has rolled out a line of modules that cover the entire signal chain, initially mainly in power products. Traditionally, Linear has sold products on a discrete component basis. These modules are complete solutions that the customer just has to solder to the board. More and more companies seem to lack the in-house skills to develop analog solutions with discrete components, so modules may be the future.
Linear’s capital expenditures, which range from $30 million to $40 million, are mainly in process technology and additional office space. This is a modest amount for a semiconductor company. It has two fabs—one in Camas, Wash., and one in Milpitas, Calif. Design wise, its finest feature is 0.6 µm, and its widest is 7 µm. Volts drive the analog product’s performance, not transistor density. Linear may go down to 0.35 µm in the future, but it has no plans to get to 45 nm anytime soon.
The company does most of its work in house, as it doesn’t feel the foundry model is suited to analog products. It feels it can improve product differentiation and control quality better in house. Linear also feels it can better control product lead times.
It does around 5% of its wafers through foundries like Taiwan Semi, and it does some outside assembly with NECB and Carsem for processes it doesn’t have in house.
As for expansion, Linear hasn’t found any large acquisitions that made strategic sense. Besides, success in integrating a large acquisition is never guaranteed. The company has completed some small technology acquisitions, though, and is always open to opportunities.
Besides the obvious impact of an economic slowdown, commoditization of the analog industry is an always-present issue. The high-performance analog products of a few years ago will tend to get commoditized over time. In addition, it is much more difficult to find analog engineers as opposed to digital engineers, particularly highly skilled analog engineers.