Programmability will win out: Interview with Nigel Toon, chief executive, XMOS

Nigel Toon, the recently appointed chief executive of Bristol based XMOS, has been involved with a number of technology companies – established and start up – over the last two decades. How does he see the changes? "The business environment is very different," he asserted. "While the market has continued to grow and mature over the last decade, the costs of developing at the leading edge have increased. Finding a niche for a new semiconductor business has become harder – but it's not impossible." Toon is well versed in new semiconductor businesses. Having left Altera as vp Europe, he was a founder of Icera Semiconductor, then moved to Picochip as it looked to enter the mainstream. Now with XMOS, he sees similarities between the three. "All are located in the Bristol area. It's one of the strongest locations in Europe for semiconductor talent – and building a team is a key part of any business. There are few places outside of Silicon Valley where you could build those types of company." Where Icera and Picochip were addressing mobile phone technology, which can be considered as a vertical market, XMOS is taking a more horizontal approach. "Building a company focused on a vertical market is challenging," Toon pointed out. "You have to identify an emerging market, then build a compelling system solution. It's not just chips; the company with the best system solution will win. You then have to win the top customers so you have a leading position before larger competitors wake up to the opportunity." He sees that approach as one of 'make or break'. "If the market grows quickly, the large investments needed can be recouped. If the market grows slowly, competitors can catch up and the return on investment can be reduced." The model has been for venture capitalists, or VCs, to fund electronics start ups. But things have changed. So how hard is it to be a fledgling company in today's electronics industry? "The environment has changed dramatically," Toon said. "Pension funds, institutions and the large 'fund of funds' investors – traditional investors into VCs – have switched their focus to the shorter term, looking for more liquidity. "VCs have therefore switched towards more capital efficient investments that generate a faster rate of return. This is, in general, not good news for the semiconductor industry because businesses take longer to grow than, say software or web ventures, and require more capital." The result, said Toon, is fewer investors. "But they are there if you know where to look." One of the perceptions of the VC business is the spectacular exit strategy, where a company goes public or is acquired for a large sum, with the investors doing very well. "Spectaculars have always been a distortion," Toon contended. "Most entrepreneurs work hard to build successful businesses and it is only when a business is working that you should think about floating on a stock exchange or merging with another company. Anyone who starts a business looking to generate a massive return will fail, because they will be focused on the wrong things. Focusing on a 'quick buck' will often mean a lack of focus on customers." Toon believes XMOS is well positioned when it comes to market opportunity and to attracting investment. "XMOS is addressing a massive market with its technology and we have the potential to build a strong, sustainable business. This is attractive to investors and, because the company is already someway down the road, the risk profile is lower and the time to return is shorter." He admits, however, that XMOS needs further funding before it gets to a break even point. "It's difficult out there," he said, "but we have supportive investors and they know we're doing something different." One common theme between the companies with which Toon has been associated is they deploy or have deployed some form of programmable technology. Does Toon think programmable technology has broken through in the way some people anticipated? "In the late 1980s, when I started at Altera, we expected plds and fpgas to replace ttl and to become part of the growing market for asics. But fpga technology carries cost and power penalties, which has limited its penetration. "When we started Icera in 2002, we believed there would be a shift toward consumer products and thought we could build a new type of dsp that would match an SoC in terms of power and performance, but which could be adapted to changing market conditions. Our first chip was 25% the size of competing solutions and delivered double the data rate. "At Picochip, we used multiprocessor programmable technology to shrink a basestation to one chip. "Today, at XMOS, we are using programmable technology to deliver new levels of performance and flexibility to the microcontroller market, while hitting price points and power levels not possible with other programmable approaches. "Horizontal companies need to use programmable technologies so they can leverage their investment across a large customer base. With a horizontal platform technology, it is the company that makes its technology easiest to use, who will win. This is what Microchip, Xilinx and Altera did and this is what we are doing here at XMOS," he concluded. Nigel has a background in electronics and technology marketing and experience of establishing, building and leading global technology businesses. Prior to joining XMOS, he was president and ceo of Picochip. Previously, he was a cofounder of Icera, where he was a member of the board of directors and led sales and marketing activities. Prior to Icera, he spent 13 years at Altera, latterly as vice president and general manger at Altera Europe.