Magma Design Automation (Nasdaq: LAVA) finally busted through the steel-reinforced doors of the IPO market in November — quite a bit later than its investors and founders may have predicted when the company was conceived in 1997. Magma, with its solid technology foundation in the EDA space, was a welcome arrival for IPO-starved investors, and the stock has held up well. Indeed, most of the tech IPO class of 2001 — companies like Riverstone Networks, NetScreen and Magma — come across as solid and respectable compared to their frothy, hype-driven cousins from 1999 and 2000. After all, these three went public in a period that emphasized profits over cool logos and clever packaging.

But though Magma seems to be on the right track, it is a company that at one time was very much a child of the high-tech bubble, and will have to prove its worth.

Founder and CEO Rajeev Madhavan started Ambit Design Systems in 1994 and was turned down by 34 Silicon Valley venture capitalists. Madhavan ran out of payroll money twice, and he had maxed-out all his credit cards. “Every morning I would walk into an investor meeting, get rejected, go back into the office and act as if we were going to get money tomorrow. Because otherwise people would bail,” he says (Ambit eventually took off, and was acquired by industry leader Cadence).

But by 1997, when Madhavan left Ambit, the tech world was in a different mood. Finding the conceptual idea for Magma’s core technology in a book by Sun Microsystems’ researcher Ivan Sutherland, Madhavan plunged back into startup mode. He called up famed technologist and angel investor Andy Bechtolsheim and asked him for a $1 million seed investment. Bechtolshiem offered him $2 million after a 35-minute conversation. Times had obviously changed.

Indeed, raising money was never a problem for Magma, which raised (and burned a large portion of) more than $112 million in venture capital to execute its ambitious plan. The amount of capital raised was a function not only of Magma’s admittedly rather liberal boom-time spending habits, but also of the product vision itself.

A New Approach

Madhavan read Sutherland’s book on a flight back to the U.S. from India — after promising himself he was done with the EDA industry. The core idea was that the estimated delay (essentially the speed) in a chip design could be formulated as a function of the gain (the ratio of the output and the input capacitances) of the design. The concept invited the possibility of a new approach to semiconductor design automation.

At the same time, a key technical issue was facing the semiconductor industry. Traditional EDA tools address the delay that comes from the gates in a chip. Gate delay accounts for more than 90 percent of delay in an outdated one-micron chip design. But as chips get smaller and more complex, signals on adjacent wires impede each other, and delay in the actual wires of the chip becomes the lead problem. As much as 80 percent of the delay in new .13-micron designs is wire delay. This is a problem that has been the topic of discussion and research in the EDA space for some time. The door was open for new ideas.

Magma borrowed from Sutherland’s conceptual work to develop an EDA product that bases itself on the absolute timing delay inherent in a chip, rather than on the area of the design (as is the case in tools built by Magma’s competitors). Executing this concept, Madhavan says, “meant re-doing the entire EDA history from scratch.”

This engineering task took up considerable resources. “This company could not have been done in today’s economic climate,” Madhavan says. “I could never have raised the money I raised.” He also admits that during the boom, Magma was sufficiently caught up in the frenzy that it spent indiscriminately and lacked the necessary structure for measured growth. Before its IPO, the company reworked its operations. “Now everybody in every part of Magma is conscious about expenditure,” says Madhavan.

In addition, Magma operates a leasing revenue model rather than a perpetual license model as is customary in EDA — resulting in a longer path to profitability and higher up-front costs in the startup phase.

Observers like Needham & Co EDA analyst Jeffrey Macy suggest that the company had to go public because it was short on cash. Madhavan says this is simply not true, citing the need for visibility and credibility in the market as the principle motivation for going public. The company raised a $25 million round of venture capital as an insurance measure just before filing for its IPO.

Fighting for Space

Madhavan is extremely bold in his predictions for the future. “To be very blunt, I’m not very highly respectful of Cadence on a competitive front at this point,” he says. “And Synopsys will be there for small blocks, but not for big designs.”

These are fighting words for a newly public company in an industry where Cadence and Synopsis are viewed as immovable pillars. “When we first came out, the response from competitors was that fixed timing couldn’t work,” he says. “This company is built to take the design leadership in EDA and be a top three EDA player very quickly.”

Needham & Co’s Macy is cautious in his assessment. “It’s definitely a new twist,” he says of Magma’s technology. “But it still needs to be proven in the marketplace.”

Turning Magma into an EDA heavyweight will clearly be a challenging task. Companies like Cadence have a massive base of legacy users that Magma will need to convert to its platform. The company’s 35-year-old CEO is confident that with his patented technology he can undo years of deep relationships between his potential customers and the top companies in EDA. It will be a market-share game. So far, Broadcom has adopted Magma as its tool of choice, and Magma has shown some initial customer success. Macy suggests that early on some customers were unhappy with Magma’s service because the company was trying to service too many clients.

Madhavan has obviously made turning Magma into a viable threat to Synopsys and Cadence his personal mission. “I am heavily competitive, there is no doubt about that. Companies like Synopsys and Cadence, by battling us with fear, uncertainty and doubt, have given me even more motivation,” he says, “This is a unique time in EDA, where we have a chance to redefine the space.”

Again, Macy is skeptical. “All of the big three EDA companies had a chance to buy this company, and they all turned it down,” he says. “Granted the valuation was probably pretty high. If they really thought that this thing had such a great dynamic future why didn’t anybody buy it? I’ve got to believe that Synopsys thinks that what they have and their strategy will be very competitive.”

Synopsys’ late 2001 acquisition of troubled EDA leader Avanti will also provide competitive problems for Magma, because it gives stability and legitimacy to Avanti, which until the recent acquisition was seen as a lost company. Avanti’s foothold in the market will be tough to dislodge.

Magma will have to prove that its new approach really is a paradigm shifting innovation, and that Synopsys and Cadence can’t come up with something comparable. Macy suggests that the industry leaders can and will come up with similar kinds of products.

But the competitive juices are flowing at Magma. Madhavan even blasts Cadence and Synopsys for being unfriendly to innovation by not allowing startups to interface with their software. He insists that Magma will change that. “We want to build a bridge to allow this to be the platform for future EDA designs,” he says.

For the young and extremely promising company, the fast and furious days when it could burn cash and run unchecked after a vision are over. Post IPO it’s time for quarter-to-quarter performance, and steady gains in market share. Madhavan will need to back up his claims and get into some key strategic accounts. From there it will be for the customers to decide. si