By KRYSTLE CHOW
Published in the Ottawa Business Journal newspaper and website.
Feb. 26, 2007
Click here to view this article on OttawaBusinessJournal.com.
At a recent intellectual property trade show for the microelectronics industry, Mosaid’s Michael Kaskowitz told the joke about the guy who calls a repairman to fix a broken radiator. The repairman pushed one button, the rad started working, and he charged the customer $100.
“But all you did was push a button!” said the customer, to which the repairman replied, “Yes, but I knew which button to push.”
Mr. Kaskowitz said the third-party intellectual property (IP) market was similar, with vendors charging clients for expertise in making a product, rather than for a material product.
Third-party IP use is making waves in the tech sector, especially in the microelectronics and semiconductor industries, where increasing complexity in chip design is forcing companies to look beyond in-house engineering.
But it’s not all smooth sailing for the fledgling trade, industry insiders say.
Third-party IP involves producing something which is then licensed to a client who will integrate it into the client’s product. It’s a growing trend, with as much as 40 per cent of a chip design today being outsourced, said Mr. Kaskowitz, a senior vice-president at Mosaid.
“Ten years ago, nobody was using IP, because engineers preferred to design the chips themselves,” he said. “However, the trend is moving towards around 70 to 80 per cent of a chip design being outsourced in the next 10 years.”
Jim Roche, acting chief executive of CMC Microsystems, said the market is maturing, and is at the same stage that the fabless semiconductor market was 15 years ago.
“There’s a fair amount of money being generated, but it’s a tiny fraction of what’s going to be earned in the next five years,” he predicted.
However, Mr. Kaskowitz took note of several trends which may become industry concerns. He called the contraction of the IP market a major development, with larger companies snapping up smaller IP vendors in attempts to expand market share across product lines.
“There used to be thousands of players in the market, but it’s now down to about 200, with only a few big players,” he said, adding that 67 per cent of the industry’s profits are held by the top-10 firms.
Mr. Kaskowitz said the trend isn’t necessarily bad, with products more advanced than ever. The market consolidation could mean more product confidence and, eventually, bigger profits for the whole industry.
“There’s a $2-billion market for IP at the moment which could grow to about $5 billion in the next five years, and about 40 to 50 product lines, each worth about $30 million to $40 million,” he said.
He said small firms are limited in how much they can grow, which also limited the market as a whole. As well, he said bigger clients are less likely to buy IP from small firms because of the risk, increasing the need for larger vendors to buy up innovative startups. That brings up the issue of IP quality, and how to ascertain value, as IP is essentially an intangible product.
“People always ask us what exactly we are selling,” Mr. Kaskowitz laughed. “We’re selling privilege, which is a hard thing for people used to hard entities. It’s not an obvious thing to assess the value of that.”
Clients tend to focus on price when comparing different sets of IP, which he likened to comparing a Mercedes to a Hyundai in the auto market.
Ken Wagner, director of design services for IP user PMC-Sierra, said it’s difficult for clients to be completely confident in the IP they license, since they can’t see whether the technology will work with their products until they physically manufacture the chips.
“Clients are stressed because they’re not in the position to look at manufacturing issues with IP. It takes a lot of partnering between vendors and clients (to build up that trust),” he said, adding there’s a “healthy skepticism” about the viability of some IP. “It’s the silicon that decides.”
Mr. Roche said problems stem less from the fact that IP workmanship is shabby, but from the fact that the methodology might not work with a client’s existing product.
“The technology may be brilliant and the quality of the engineering excellent . . . but all it means it that they’re having problems integrating the technology, and the methodology isn’t as mature,” he said.
He compared the mismatch to Stephen King using a ghostwriter to write one chapter in one of his novels – even if the writer was Margaret Atwood.
“It’s not that the writing is bad . . . it’s a different level of artistry,” he said. “It doesn’t fit into the complicated set of processes that’s already been established.”
It gets even more complicated with mixed-signal IP, where chipmakers mix analog and digital technologies on a chip.
“It’s difficult to do. There are only about 5,000 to 10,000 engineers worldwide who can work with analog signals,” said Mr. Kaskowitz.
Ultimately, however, the big issue with IP is deciding when to outsource know-how and then building the trust between vendor and client.
“(That) scares people – it’s like with cars where people can’t go out for a drive by themselves without knowing where the gas station is, because they can’t fix the car themselves if it breaks down. It’s too complex,” Mr. Kaskowitz said.
As such, it’s often not just about the technology, he explained.
“The technology has to be good, but the relationship (between clients and vendors) has to be flawless,” he said. “It’s like marrying them.”