The road to success is long and hard for eda start ups

4 mins read

EDA and IP were common themes amongst the three start up companies featured at this year's Sophia Antipolis Microelectronics Forum.

Each had a valuable opportunity to present their ideas and business plans and to secure one on one discussions with major players, including ARM, Cadence, Intel, ST and Samsung. Synflow from Brittany, introduced Synflow Studio, an integrated development environment that sits midway between Verilog/Vhdl and C based high level design. The two man company is looking to 'change the way that electronics hardware is designed'. Based in Montpellier, Algodone is focused on the IP market and aspiring to the concept of a universally interoperable 'IPtunes' library. Its first step is a software tool designed to help IP vendors package their IP blocks for secure (encrypted) interoperability. Meanwhile, Lyon based Zettice is straddling the IP and design sectors, with IP blocks and an IP optimisation service based on its own tools. It already offers more than 40 arithmetic IP blocks, with another 50 complex data/signal processing IP algorithms underway. Aiming high, one target application is financial risk analysis. One common thread between these start ups is that they are run by young enthusiasts who believe they can make a difference. And while their longer toothed peers may take a more jaundiced view of their prospects, most welcome and encourage such companies. In the past, such start ups have been acquired quickly by larger companies, but it's not as common as it was. So have the big companies changed their strategies? Wally Rhines, Mentor Graphics' ceo, said: "Start ups are important to the continued growth and vitality of eda. Most of the revenue growth in eda comes from developing the capabilities to solve new problems and start ups can often zero in on a newly perceived challenge and focus completely on addressing it." Synopsys, meanwhile, is more cynical, as John Chilton, general manager, strategy and business development, pointed out: "The problems of designing today's multibillion transistor chips are huge and systemic. It's increasingly hard for a small group to solve these problems in a meaningful way." Agreeing in part, Xilinx' EMEA marketing director Giles Peckham added: "Tasks such as synthesis or place and route are now on the critical path to silicon and we have to 'own' this technology, not partner with third parties." Xilinx has recently completed a 1000 person-year ($200million) development of its Vivado fpga design suite, replacing a 15 year old tool acquired with then start up Neocad. "However, there is more room for specialist third party tools," Peckham emphasised. Mentor also recognises the eda industry is composed of many niches representing different technologies, methodologies and tools – each aimed at solving a specific design problem. "No company can offer solutions in all of these niches," Rhines suggested. "So Mentor believes it is essential to be open and to support customers that need to integrate solutions from other providers, many of whom are small emerging companies." Support and encouragement In addition to its standard Open Door program for eda tool partners, Mentor offers a Term for Start-up programme (TFS), tailored to the needs of start ups. "Its a low risk approach, allowing start ups to purchase what they need at their current stage of development at an attractive price and with simple contract terms," Rhines explained. "We recognise that a start up today may be a large customer tomorrow – if given a little help." To encourage partners, including start ups, Xilinx' Alliance programme operates at three levels. Synopsys' scheme, called Insync, gives start ups access to its tools to demonstrate interoperability. Support for eda start ups is also available from customers. STMicroelectronics, for example, is seen as a useful testbed for new SoC design tools, technologies and methodologies. But, as with many of the major oems focused on complex SoC design, finding the right person, running the right project at the right time is the challenge. However, an endorsement from a company like ST can make a huge difference. For example, after 10 years in business, US based eda company Carbon Design Systems entered into a strategic partnership with Samsung, which claims that Carbon's technology has helped it to develop leading edge consumer devices. Interestingly, it also received $4m from Samsung Venture Investment. Exit strategies A common failing amongst start ups, as business angels point out, is insufficient attention to the business plan and the investment required, not only for product development, but also to sustain the company. In the eda sector, the problem is acute, as Chilton pointed out: "Investors are not putting much money into eda startups." A contributing factor, says Chilton, is a lack of really profitable exits in eda. Fewer eda companies are reaching the point where they can go public and it is taking much longer for those that do. Companies have to be much bigger and more profitable, which requires very patient investors. That means acquisition is a more common exit strategy than IPO. An outstanding technology proven with real customers may prove sufficient motivation for the acquirer; the start up may not even need to be profitable. Xilinx keeps a close eye on specialist technologies and will make acquisitions when they become strategically important or represent a high potential market area. For example, it came across a French start up offering its customers a particularly effective way of reducing dynamic power consumption. "We acquired the company as we felt it would give us a competitive advantage," Peckham said. High level synthesis is another area where Xilinx sees strong potential. "There were two good start ups in that market. Synopsys acquired one and we did not want to leave the other to our competitors," Peckham remarked. Synopsys will typically acquire several start ups per year. "If their technology is compelling and they would like to join us," Chilton qualified. "But the bar is high; we like to see that people have proven their ideas in the real world with real customers and those customers have been successful with real designs." Clearly, Chilton has experience with start ups. "It is easy to describe an idea; it's another thing to make it work in the pressure cooker of a tough tape out."