Innovative Ideas for Predictable Success
      Volume 1, Issue 4

  NEWS  |   CALENDAR  |   PAST ISSUES SYNOPSYS.COM  |  CONTACT US


  Customer Spotlight
Spotlight picoChip: From Startup to Success
Rupert Baines, Vice President of Marketing with picoChip, offers sound advice for would-be IC startups considering the venture capital route to funding, and recounts picoChip’s early experiences on the road to success.

Back in 2000, two individuals, Peter Claydon and Doug Pulley, brought together their respective expertise in the semiconductor and wireless domains to develop a product for a specific market opportunity. Their shared vision was to provide a solution to the challenges of the next generation of wireless systems: cost, development time and flexibility.

In the best tradition of technology startups, the first nine months of the company’s history were spent working from one of the founder’s kitchen tables while the business plan was refined, the management team was recruited and VCs were courted. picoChip was bootstrapped from the co-founders’ and early employees’ own funds, working for a minimum wage until A-round funding was secured from technology specialists Pond Venture Partners and Atlas Venture.

More Than Just Capital
“Getting backing from the right VCs is critical,” explains Baines. “Backing is the key word. It’s not just the level of financial investment that will take you to that key milestone, it’s important that there’s also a deep level of belief and trust that will see you through the tough times ahead. The euphoria of getting the term sheet signed off and the cheque in the bank doesn’t last that long. It’s really important that you are mentally prepared for the inevitable difficult times that lie ahead. You want your VC to be right in there with you.”

Besides the finance, Baines believes that the right VC will provide the benefit of experience, strategic counsel, long-term commitment and invaluable connections within the industry, further emphasizing the advantages of working with a technology specialist investor.

Baines feels that there is still a lot of variability in the industry in terms of just how technology-savvy investors are. “You don’t necessarily have to deal with semiconductor industry veterans, although there are an increasing number of those people involved,” he explains. “There are plenty of people out there with accounting-type backgrounds who really do understand the technical issues that can have a major impact on the startup’s business – that’s what really counts. They may not be engineers by training but they are shrewd and experienced business people who understand what’s important and where the pitfalls might be.”

Assessing the Risks
The technology-savvy VC will take a long hard look at your business plan in terms of the risks associated with the product development. Part of the technical due diligence process will seek to understand the parameters of the proposed development schedule and technical complexity. “A good VC will ask probing questions about your development plan,” claims Baines. “One of the things that has changed in the UK investment scene over the past five years is that we now do have a critical mass of VCs who are experienced and understand the broad issues associated with chip design, and that has to be a good thing.”

Baines agrees that for startups like picoChip, you are literally betting the company on the successful outcome of the chip design. “It’s a bit like a parachute jump,” states Baines. “Only on the way up in the plane you’re designing your canopy, and sometimes you’re still strapping it on when you’ve already jumped out of the plane.” What’s more, this isn’t just a one-off. Assuming you land safely, it’s back in the plane for another jump – a process that puts the development team under tremendous pressure.

The cost of building an advanced chip is well beyond $10m, arguably higher for advanced geometries of 90nm and below. There are not many startups that will be able to afford that level of investment in a first project, and so clever cost reduction is often a key factor in achieving early success. However, this must be done without increasing overall risk. By definition, a startup will be trying something new and innovative, which in itself increases the risk associated with the project.

“It’s called venture capital for good reason,” continues Baines. “You are trying to do something nobody has done before, so it is going to be risky. But that doesn’t mean you take all of the risks at the same time. At picoChip, our innovation is fundamentally within the architecture – that’s where our competitive advantage lies. Because of this, we do everything we can to de-risk the chip implementation.”

According to Baines, picoChip spends a considerable time analyzing exactly where the risks can be managed. This includes taking a conservative approach to engineering the chip, which means, where possible, using a generic process and working well within the design rules, and not pushing the limits on clock speed. With its most recent tapeout, picoChip has hedged its bets by processing the design in multiple lots. The first is rushing through on a fast shuttle so they can get their hands on silicon as soon as possible, but by way of an insurance policy they have a second lot being manufactured on a seven process corner skew. This approach will provide six backup samples which cover a spectrum of process variations to provide contingency. Furthermore, half of the slow batch will be held at contact until the first silicon has been tested. This provides a further opportunity to make a final adjustment to the chip before samples are delivered.

picoChip views the partnership with Synopsys as a critical element of its overall risk management strategy. “Just as we were able to draw on the experience of our investors, it is critical that you can take advantage of the wealth of expertise from your design tool partner. Synopsys provides us with leading technology to enable us to do the design and, equally importantly, they are a rich source of expertise when it comes to achieving a predictable path to tapeout. That includes everything from front-end design and verification, right through to links to manufacturing.”

Managing Partnerships
Selecting and managing partnerships is a core skill, according to Baines. He expects the selection process, and subsequently the ongoing management, to require a significant amount of effort in order to profit from the relationship. In selecting an EDA partner, picoChip made a major commitment to dedicating valuable resource to an evaluation. “Before we placed an order with Synopsys, we spent around six months evaluating, marking, giving them test designs, test layouts, and just working with the teams so that we could see how it would work both technically and on a human level. Both parties are making a commitment and we have to view it as a long-term relationship. We want a level of commitment from our partner that will see us through a number of projects.”

So far, that commitment has paid off. picoChip’s first design, the PC101, worked first time. Largely intended as a proof of concept, getting this design right was critical in securing customer wins and B-round financing. Their next design, the highly-optimized PC102 is in volume production and deployed in basestation designs with around 50 different customers, targeting very demanding next-generation wireless infrastructure applications and delivering ultra-high reliability. Since then, picoChip has closed a C-round of funding which has provided working capital to further expand the business.

“Commercially, Synopsys demonstrated that they clearly understood the financial framework that a startup operates in. With big-ticket software licenses there is always flexibility for negotiation, and Synopsys were sympathetic to our situation in terms of the technical risks we were taking. From our side, if the PC101 had not worked, we wouldn’t be here now.”

What would be Baines’ best advice for would-be startups? “Don’t compromise. You will need the best people throughout your design team, management team and board. You also need the best advisors, investors and technology and tools partners. A lot of startups still make the mistake that thinking great technology is enough to be successful. It’s not, and you need to invest in marketing, sales and financial management, and foster close partnerships with your key suppliers. I think that your customers will notice this as well. Ultimately, something that’s fundamental to success of the business is, ‘Will the chip work?’ You are betting the company on that one chip, and if something goes wrong with the tools, well, that’s game over.”


©2010 Synopsys, Inc. Synopsys and the Synopsys logo are registered trademarks of Synopsys, Inc. All other company and product names mentioned herein may be trademarks or registered trademarks of their respective owners and should be treated as such.


Having read this article, will you take a moment to let us know how informative the article was to you.
Exceptionally informative (I emailed the article to a friend)
Very informative
Informative
Somewhat informative
Not at all informative

Register Buttom

Email this article
About Rupert Baines
Rupert Baines
Vice President of Marketing, picoChip

Rupert has spent eighteen years in technology marketing, working on commercialising many leading edge technologies. He spent six years with Analog Devices in Boston, working on semiconductors for GSM and wireless infrastructure before managing their Broadband product line and holding responsibility for the industry-standard ADSL chipset. He has also worked for operators, most recently as Director of Product Development and Strategy for Atlantic Telecom, where he ran a large pan-European DSL rollout and trialled wireless broadband deployments.
Horizontal
  WEB LINKS

-   Picochip
Horizontal
“A lot of startups still make the mistake that thinking great technology is enough to be successful. It’s not, and you need to invest in marketing, sales and financial management, and foster close partnerships with your key suppliers.”
Rupert Baines, picoChip.