by James Montgomery, News Editor
Chris Moore, CEO of metrology equipment supplier Philips AMS (now known simply as AMS — Advanced Metrology Systems) talks with WaferNEWS about the decision to break free from former parent Royal Philips, the new reality of semiconductor market cycles, and where the company is pushing new inroads for its trench-measurement technology.
Late last week, Royal Philips announced the sale of its majority ownership in metrology equipment unit Philips AMS to JHW Greentree Capital, an affiliate of J.H. Whitney & Co. of New Canaan, CT. According to Chris Moore, CEO of the newly renamed Advanced Metrology Systems LLC (AMS), both the business and its former parent (which will retain a 19.9% stake) long understood a divestiture was in the cards, as Philips narrowed its core focus to markets such as medical and lighting.
“There was never an issue with them investing in us to keep us going, and growing at a certain level,” he told WaferNEWS, “but there was no long-term home for us. It was going to happen at some point.” In fact, this deal was put together before the Philips Electronics’ semiconductor business (since renamed NXP) was sold months ago to private equity investors led by the Blackstone Group, he said.
Moore acknowledged that AMS turned down other proposed combinations with bigger metrology firms (“we didn’t want to be part of one of the big players that would gobble us up”) and smaller ones (Royal Philips would still have been the majority owner, putting them back at square one of having a foothold in a noncore market). “Private equity was a good way to get us the money we needed to grow, and get a parent to support us,” he said.
With new private-equity ownership, Moore sees a symbiotic relationship where the technology know-how is handled at the management level, while the board focuses on market strategies and growth, with a general knowledge about the business of the industry and just a “comfort level” of technical understanding. “That can stop you from getting tunnel vision, which is very easy in this industry…it forces you to look at the big picture,” he said. The private-equity model also takes a longer-term view of business growth, whereas VC or publicly traded companies are ultimately driven by quarterly results, and are tough to convince that a return-on-investment could take 18 months, if it happens at all.
In addition, life as a public company is increasingly risky and expensive — see the laundry list of companies being investigated for stock options manipulations, which is already amounting to hundreds of millions of dollars in restatements and charges. “I saw a pitch a couple of weeks ago: ‘Come to the UK, where reporting is easier,'” Moore quipped.
AMS’ 50% Y-Y growth was a big draw for equity investors, as was the company’s business strategy — “finding key enabling metrology points in new processes that companies are bringing up, then solving that particular need,” Moore explained. “String enough of those together, that’s a good business model. Not, ‘Gee, I want 10% of a big market’ — that’s not easy.” He pointed to AMS’ work with Infineon and its former memory chip unit, Qimonda, to solve a problem with etch trench structures, which led to AMS licensing MKS Instruments’ model-based infrared reflectrometry (MBIR) thin-film metrology technology in Jan. 2005. “Now we’re the dominant metrology [supplier] for measuring high-aspect ratios in trench DRAMs,” Moore said, noting the company also is working with Asian DRAM fabs including Winbond, SMIC, and Inotera.
Moore also hinted that AMS is now working with some customers to solve problems involving trench measurements for power devices — similar to what it’s done with DRAM, but with adaptations in technology and modeling.
One reason that private equity investors have become keen on the semiconductor industry is a perceived smoothing of historically cyclical market ups and downs. Moore pointed out that in the past, volatile cycles were usually driven by economic conditions and what happened to a particular end-product (e.g., PCs), but now there’s a wider range of end markets, and a down cycle in one chip segment (e.g. memory) may be mitigated by growth in another (e.g. logic). He also pointed out that equipment makers can ramp up much more quickly than in the past, and can run lines much harder to keep utilization rates high instead of reinvesting in more capacity. “Now fabs are running 110%-120% of what they were designed to do, because the equipment can let them do that,” Moore said.
Moore echoed comments made by analyst Stuart Muter at a recent SEMI breakfast near Boston, who suggested that the industry has now basically compressed its market cycles to one-year intervals — with strong orders in the first half of the year, followed by a slowdown to digest capacity, then ramping up toward the end of the year to meet holiday demand. “When it took much longer to ramp, people tended to overbuy, which made the whole cycle worse,” Moore noted. However, he pointed out that R&D cycles have actually lengthened out — e.g., low-k, high-k, and metal gates have all been shifted out beyond their originally intended nodes, as both IDMs and equipment suppliers struggle to get the technologies to work properly. “You’ve got two cycles in the equipment industry: one shorter, and one longer,” Moore said. “What you’re seeing is people getting nodes out on time in terms of size, but pushing technologies further out” — e.g., DRAM makers are not doing things at the 50nm node that they thought they’d have to two years ago, he said. — J.M.