Feature Story

More feature stories by year:


Return to: 2007 Feature Stories


APRIL 24, 2007: SOCcentral


The semiconductor industry has always grown on waves of innovation, and much of this innovation has come from fabless start-ups. The top ten fabless semiconductor companies by revenue got to their positions by pioneering in the fields of communications, graphics, flash storage and programmable logic, including such names as Qualcomm, Broadcom, SanDisk, and Xilinx.

Prosperous fabless companies have, in turn, driven the growth of other segments of the semiconductor industry, namely foundries, EDA and IP vendors. The industry, therefore, looks for a continual supply of start-ups and the opportunities they create.

But there is a big problem: It takes approximately $25M of funding to catapult a fabless semiconductor start-up to break even, and for every 10 start-ups, three will fail, four will be the "living dead," and just three will generate a return on their investment. Those that generate a return may not do so until four or five years have elapsed, so many investors have focused their funds elsewhere.

Experienced venture capitalist (VC) firms - like those on famed Sand Hill Road in the Silicon Valley - have stayed the course, but what can the semiconductor industry itself do to help matters?

The old rules for tech investment were to have a good idea backed by an experienced team. During the frenetic period of "irrational exuberance," investment criteria were relaxed, and due diligence was sidelined. With many companies receiving funding without a coherent business model, and markets being flooded with more start-ups than they could support, it was inevitable that the bubble would burst.

Post-bubble, some VCs are now looking for entrepreneurs with not only a track record, but also a product, and preferably a paying customer. In that case, it is better to get a bank loan than to give away equity! Thankfully for the start-ups, not all VCs are at this extreme.

The amount of money required to create a fabless start-up is large; it takes time to get to the first product and, hence, first revenue. Good VCs are looking for a unique, sustainable proposition that they can catapult into a high-growth trajectory.

Given these risks, most VCs expect a start-up to have a prototype and some customer interest, e.g., a letter of intent, before they will consider it for investment. But an entrepreneur requires some level of seed-stage funding to write a business plan, register the company, develop a prototype and prepare some basic patents for later filing. Without these fundamental first steps, an entrepreneur will not be in a position to court potential customers.

This gap between a VC's requirements and an entrepreneur's needs is referred to as the "funding gap," which occurs between the conceptual phase and round-one funding, i.e., the seed stage.

Government grants and business angels had typically helped fill this gap. The grants, however, are now more restricted, and there are currently few angels with the stomach to invest, given their experiences of the post-bubble rout, where angels were diluted out of their investments. Things are slowly improving, but what can entrepreneurs do to get their start-up off the ground? Well, there are the usual rules, namely:

Proposition - You need a differentiated value proposition. For every company with an original idea that gets funded, several copycats are also funded. Being lower cost than competitors is not enough; you need a sustainable competitive advantage to stay several months ahead of your competitors. You also need to develop a prototype or concept demonstrator to snag a customer. Are you servicing a global market? Don't talk about multi-billion dollar markets in your elevator pitch. Identify a niche with a "pain point" to solve and explain how you will service it, generate revenue, and deal with the competition. Many applications are supported by well-defined standards. Do you support those? How do you add value in order to differentiate and effectively compete?

Investment - Get an investor who has a track record of investing in your sector and will be in it for the long haul. Speak to the management of their portfolio companies. And dump the flashy car and travel cattle-class. Profligate entrepreneurs scare investors.

Operations - Foster a nimble start-up culture that's responsive to markets and customers. Keep expenses down and conserve cash. Work out the minimum salary you can survive on. Experience is key. Good talent is always hard to find, whether boom or bust; however, with many companies undertaking restructuring, a lot of experienced "grey hairs" are on the market.

Focus on what you're good at and partner where it makes sense. Pick a credible partner; don't be afraid to ask for references. VCs will look for a strategy that includes offshore and outsourcing elements (verification and software development are possible areas).

In the investment proposal, keep capital expenditures to a realistic minimum. Manage your suppliers, e.g., cut a deal with an EDA vendor. They are in it for the long run, so they should be able to lower upfront costs; they have a stake in your long-term success.

Exit - Plan for delivery before planning for an exit. There have been few fabless IPOs in the last five years, and the preferred exit route is a merger-acquisition. What makes yours an attractive prospect will be a product line with paying customers, and a roadmap to more of the same.

Those are some of the basic tips. But what about development? This is the bit you are supposed to know best, but how does that fit with what VCs want?

VCs want start-ups to create multiple products that are also differentiated, which - if they are to be sold globally - must support multiple standards. They must also make money, which means they must be developed within budget and released within a given market window.

This is a challenge for even the most cash-rich company. Design windows are continually narrowing, and cost pressure is driving intensive integration. This, in turn, drives complexity - resulting in 80% of engineering time being spent on verification.

It is, therefore, vital to reduce the overall business risk. Two major contributing factors to that risk are development risk and time-to-market. Given that many high-growth applications are supported by technical standards, which allow differentiation, it would be of great value if entrepreneurs could gain access to "platform" technology to which they could add their key differentiator, or "secret sauce."

Many intellectual property companies — including Imagination Technologies — now offer innovative, or even unique, programs that allow entrepreneurs to license a complete IP platform, allowing the speedy development of a prototype to demonstrate to prospective investors and customers. With these programs, an entrepreneur might only pay for the license fee when they're ready to tape-out, by which time they've already secured first- and second-round funding, or they may pay for part of the license in equity.

Programs like these reduce development time and risk, thereby improving time-to-market. All these factors reduce the overall business risk for the entrepreneur and investment risk for the VC.

Return to: 2007 Feature Stories