On the morning of September 11, LogicVision CEO Vinod Agarwal was in a hotel room in Boston, waiting for morning meetings with investors, as part of LogicVision’s IPO road show. The next day he would be moving on to New York, for a meeting in the World Trade Center. None of the meetings took place.

“All of a sudden, you can’t deliver, and it’s not in your control,” Agarwal reflects, remembering his feelings that day. “You’re the CEO of a company and you’re going on the road with your CFO, and all of a sudden something of such dramatic proportions happens. You really have to ask what your priorities are. And yet you still have to run a company.”

Agarwal and his CFO were trapped in Boston until airports re-opened. He admits that running LogicVision was not foremost in his mind, and that he was glued to the TV.

When he was finally able to return to Silicon Valley, it was to pick up the pieces of the abruptly terminated road show, and start moving forward again. “You question things,” he remembers. “You’ve spent so much money on the process, and everybody’s morale in the company is low — you were saying that you were going to go public, and then all of a sudden, this.”

Technology Shift

The canceled road show came on the heels of nearly a decade out of the spotlight for LogicVision. The company’s consistently low profile was in part due to traditional venture-capitalists’ and investment bankers’ lack of interest in the company and its “embedded test” technology for the semiconductor industry.

The basic concept for LogicVision grew out of Agarwal’s masters thesis. But it was only after 14 years as a professor at McGill University in Montreal and a consultant to Northern Telecom (now Nortel Networks), that he started the company in 1992 -- on $3 million from Canadian venture investors.

The late 1990s boom times were yet to begin in the technology industry. LogicVision was founded on the kind of technology leap of faith that has characterized technology innovation from the early days of Silicon Valley.

The semiconductor industry spends billions of dollars testing the chips it produces for defects. This testing portion of the chip design cycle has traditionally been totally reliant on huge testing machines called ATEs (automated test equipment), made by companies like Terdayne, Advantest, and Agilent.

ATEs cost upwards of $5 million a piece. On top of their hefty price tag, they are the only piece of the chip manufacturing and design process that isn’t agnostic to the type and size of chip being produced. In other words, manufacturers end up having to buy huge testers for many of the new chips that they roll off the production line. For the semiconductor industry, this translates into billions of dollars spent annually on testers.

The concept behind LogicVision was to embed a tester in each chip, so chips could be tested from within, rather than having to be tested by an ATE. This could potentially save a tremendous amount of money for semiconductor companies, by circumventing some of the need for large and costly testing machines.

“We’re commoditizing the test operation so that it doesn’t matter what kind of chips are coming through,” says VP of marketing Mukesh Mowji. LogicVision was the first commercial entity to propose embedded test technology, and it is now, according to Mowji, the “gorilla” in the space.

That dominance doesn’t yet translate into billions of dollars, however — LogicVision had revenues of $5.4 million in the quarter ended December 31. But Agarwal is confident that as chips get more complex across the board, the embedded test market will start to look more like the current market for ATE testing.

In fact, Agarwal talks about embedded testing eventually replacing ATE as the de facto test solution for complex chips in the semiconductor industry. He expects the total market to be about $1 billion in five years — or about one third of what semiconductor companies would otherwise have to spend on traditional testing solutions.

The transition can’t happen overnight, since it involves a totally new way of doing things, and changes the dynamics of the market. The vast majority of chips are still tested using ATEs — and that will continue to be the case, especially for less complex designs.
“It’s the difference between microcomputers and mainframes,” says Agarwal. “There was no way to take what mainframes were doing and turn them into a little chip — you had to think differently.” Making semiconductor firms think differently is obviously the key task ahead for LogicVision.

The company already has more than 100 customers ¾ including giants like Cisco, Intel, Texas Instruments and National Semiconductor — who use LogicVision’s technology mostly for complex, one million gate-plus System on a Chip (SOC) designs used in workstations, networking equipment and complex consumer handheld devices like mobile phones or PDAs.

Agarwal and his team are expecting not only to take away market share from existing ATE players, but also to alter the functionality of semiconductors in general. Since the tester is in the chip, semiconductors with embedded test will able to self-diagnose malfunctions in the field. This makes embedded test not only a potential replacement for traditional testing in the de-bugging and manufacturing phase, but also adds a whole new dimension of capabilities that Agarwal deems important for the devices of the future.

ATE vendors Teradyne and Credence seem to have acknowledged that embedded test is likely to grow by actually investing money in LogicVision, alongside Intel.

“Tester companies are doing what is right for them,” says Agarwal. “They recognize that the change is coming and the customers are asking them to make a change. And they are doing what is in their interest, which is to work with the leader in this space. And it’s in our interest to work with the tester companies because they understand the needs of the customers.”

Of course, a company like Terdayne with $3 billion in revenues still dwarfs LogicVision. Agarwal has some catching up to do.

The Road (Show) Less Traveled

By most measures, Logic Vision is doing well. It turned a pro-forma profit last quarter, something that a lot of technology companies can only hope for at the moment.

Five weeks after the September 11 attacks halted the company’s efforts to go public, Agarwal and his CFO found themselves back in Boston, in the same hotel room, ready to get the process back on track. And this time they went through with it. The company went public on October 31 ¾ the first technology IPO after September 11. Agarwal recently did his first earnings call.

But the Logic Vision has, in a sense, been ready to go public for some time, with a full-service management team in place for several years. Mowji jokes that the testing portion of the semiconductor design cycle is “that step-brother nobody wants to talk about,” when in fact it’s a massive market. The front-end design phase — dominated by giant EDA players like Cadence and Synopsis, gets much more attention.

When dozens of technology companies were hitting IPO jackpots, LogicVision was put on the back burner. “Even though we had a technology that will make a major difference in the semiconductor business, the bankers were positioned after dot-coms,” Agarwal explains. “We weren’t sexy, we were real and fundamental, there was no makeup on it.” So as companies all around hit the public markets, LogicVision had to stand and watch.

“You question yourself. Your next-door neighbor becomes a hundred millionaire and you’re still working day in day out and you wonder what happened,” says Agarwal.
Only when the bubble burst, and IPO underwriters became desperate for companies with business fundamentals, did LogicVision start to get attention. By then the company was also farther along in terms of proving that embedded testing was a lasting reality.

Why Now?

Agarwal admits that he chose to go public at the worst possible time, and investors constantly questioned his timing during his road show.

In the end, the company sold 585,000 shares at $9 per share, for total proceeds of approximately $45.8 million — a modest total for a company with such big plans.

But Agarwal explains, “Our customers are taking big risks in making a transition to a new technology, and they want to make sure that they are working with a company that has financial transparency, viability, growth, etc.” In other words, to really promote embedded test as a serious technology paradigm of the future, LogicVision had to become a public company.

As it stands, LogicVision is yet to garner a lot of attention, or analyst coverage. It will probably have to contend with limited visibility until it starts to show bigger revenues and the embedded test market really grows. Competition will emerge with the market — from established companies as well as startups — and LogicVision will have to maintain its current category-leading position. EDA players Synopsis and Mentor Graphics are already entering into LogicVision’s market.

The semiconductor industry has seen better times, but Agarwal feels that the industry’s slump will actually accelerate the adoption of embedded test, since CEOs won’t have the budgets to buy big testers.

LogicVision’s stock (Nasdaq: LGVN) is maintaining its value, and Agarwal checks it once or twice a day. Growing the company will be a challenge for Agarwal, but at least he has the satisfaction of finally being at the helm of a public company, roughly a decade after leaving academia to become an entrepreneur.
si