Mark Downing, vp strategy, Silicon Laboratories

SiLabs reported sales of $104.5million for the second quarter of 2009, 25% up on the previous quarter and with a margin of 62.2%. "It was a good rebound from Q1," said Mark Downing, vp strategy, who added: "We have outperformed the competition." Interestingly, design wins were said to be up across the board. SiLabs has undergone a degree of change during the last few years. "The company is diversifying," Downing continued. "A couple of years ago, we divested our cellular business to NXP. It was a tough decision, but we've managed to replace the $180million of lost revenue with more diversified products and with higher margins." In 2004, 50% of SiLabs' business was handset related. Today, it's nearer 15%. SiLabs has around 14,000 customers and the top 10 includes such market leaders as Samsung, which uses SiLabs products in handsets, set top boxes and tvs. In Downing's opinion: "The good thing about Samsung is that it's not only receptive to new technology, it's also a significant player in a number of markets." SiLabs' approach is focused on the design of differentiated products based on its mixed signal technology expertise. "It's our mixed signal integration abilities that differentiate us from the competition," Downing contended. "This translates into smaller footprints, programmable embedded mcus and reduced bills of material. All of these factors drive the adoption of our products." One reason why Downing thinks SiLabs has a strong future is the consequences of Moore's Law. "Digital circuits scale well with process shrinks; analogue circuits don't," he claimed. "As products go to 45nm, we'll see analogue splitting off and that will need a move to a system in package approach. We're now shipping tens of millions of SiPs without too much cost disadvantage. Ten years ago, SiP was talked about, but it wasn't cost effective. Today, it is and I can see opportunities in applications such as metering, where there's a need for a radio and a low power mcu." He said the combination of reduced voltage swing tolerated on these smaller geometry processes and the noise isolation challenges seriously impairs the parametric performance of the integrated SoC device. "I think this will drive a new paradigm, where the SoC is purely a digital chip (processor plus memory plus logic) and the analogue/mixed signal functions will be integrated on a companion chip." And Downing thinks this companion chip will, in many cases, be programmable and may integrate either a dsp or microcontroller, with firmware addressing specific application or customers needs. "SiLabs would be well positioned to address this mixed signal companion ic and therefore be a strong beneficiary. In my opinion, this approach makes a lot of sense from a system cost and performance perspective since you can optimise the choice of process technology," SiLabs believes that differentiation in established markets means market share can be gained quickly. "For example, our fm radio business has grown by 20% a year over the last three years and will grow again this year," Downing observed. SiLabs' approach is also based on generating high margins and the latest financial figures showed margins at more than 60%. "It means we can continue to invest more than 20% of sales revenue in R&D each year," Downing claimed. "The industry has high growth potential and we see opportunities to drive growth with differentiated products, maintaining our overall growth of 20%." Despite this high level of R&D investment, SiLabs focuses much more on the D aspect than the R. "But we do some development around memories to ensure we maintain a competitive market position," said Downing. So where does the R now get done? "VCs have been helpful in taking technology and developing products. Companies like SiLabs benefit from acquiring them. But it's hard for VCs to get a return on their investment at the moment and this may create a void," Downing continued. That requires investment in other companies or their acquisition. "We've recently invested in a company developing MEMS based resonators. That was a strategic move for our timing business. The company was looking for money to develop its technology and we're looking to take the technology into mainstream foundries and to market. But these kinds of opportunity may become fewer. "We're still looking for acquisitions that provide an immediate synergy and for companies in adjacent markets. The recent investment in the timing company was an opportunity to bring new technology into the company and to expand the TAM for timing products," he said. Even so, Downing claimed SiLabs was doing new product development in a 'disruptive' way. "This is bringing significant benefit to customers," he claimed. "But what's also helping to drive growth is that we've been able to maintain our r&d spend. We've taken a deliberate position to maintain the program. We've also got a strong focus on cost reduction. By designing chips for the smallest possible die size, we're lowering the customer's BoM," he concluded.