Two players have dominated key segments of the semiconductor lithography equipment market for years.
TEL controls a huge portion of the track-tool sector, while Cymer holds the lion's share of the laser light source marketplace. But the two companies' grips on their respective segments have loosened a bit, with the emergence of Sokudo, the Dainippon Screen/Applied Materials track-tool joint venture, and Gigaphoton, the Komatsu/Ushio-backed light source manufacturer.
During a one-on-many briefing at the SPIE Advanced Lithography event, I had a chance to meet with several key members of the Gigaphoton team, including Yuji Watanabe and Hakaru Mizoguchi, the company's president and corporate executive VP, respectively. This was the first time the execs had come to the U.S. to talk with the trade press, and as it turned out, I was the first American editor they met with. One of their managers, Olivier Semprez, delivered most of the presentation, with Watanabe-san and Mizoguchi-san adding to the conversation. Semprez is a Frenchman, so before he began his spiel, I said a few things to him in his native tongue, which elicited laughter from the execs and helped loosen things up.
Since its founding in 2000, Gigaphoton claims to have made great strides in battling Cymer's light source hegemony. Semprez said it has averaged better than 30% growth each year since 2002, with fiscal 2006 (which ends next week) expected to be even better. The company claims 2006 market share in the 30-35% range, based on the number of units sold. Of course, if you look at installed base and historical sales, that number would come way down, but it is still impressive, in light of Cymer's strong competitive position.
Giga expects to hit the 500 mark in its own worldwide installed base by the end of 1Q07 (the end of June). Most of those laser boxes are hooked up in Japan, followed by a significant base in Taiwan, Korea, Singapore, and China, and most are KrF-type tools (although ArF sales are trending up). With marginal shares in the U.S. and European markets, the Giga execs see ample opportunities for growth in these underpenetrated regions.
Semprez and Watanabe-san explained what Giga sees as its key differentiators: more of a production- and reliability-centric R&D approach, and a focus on not just the laser itself, but on the litho cell as a whole. They cited extensive analyses of the costs of unproductive laser downtime---the cost of not operating the laser---something they said had not been previously considered.
Many hidden productivity killers emerged during the studies, beyond whether the laser was up and running or not. Even a few minutes of unscheduled automatic recalibrations, for example, can add up to many lost hours over the course of a year and millions of dollars in lost revenue, especially in a high-volume 300-mm fab. Giga wants to cut the laser's cost of ownership in half by the end of 2008, by pushing down both the costs of consumables and of downtime. They also noted the joint support agreement between ASML and Giga, with its goal of "zero wait time," as an example of seeing the laser as part of the bigger litho-cell picture.
The Giga team pointed out successes in the areas of advanced ArF tooling and EUV research. They touted the GT61A ArF excimer laser box, which can support immersion scanners with >1.3 NA and supposedly beat Cymer to market last year as the first volume production laser of its kind to be shipped.
In the area of extreme ultraviolet litho, Mizoguchi-san said they are working with the Japanese EUVA consortium, with lasers installed on both Nikon and Canon platforms. He went over the advantages and disadvantages of the two EUV source approaches, laser-produced plasma (LPP) and discharge-produced plasma (DPP). He said each face significant challenges to reaching production worthiness on a timely basis. Mizoguchi also noted recent encouraging research results, on both the LPP power enhancement and CO2 laser debris mitigation fronts. He sees DPP as an intermediate step, with LPP as the ultimate choice, especially because of its scalability, but that it is too soon to make a final decision on DPP or LPP.
Like sportsmen and women the world over, corporate execs give lip service to the need for good competitors, claiming that it gives their own team's efforts more focus, drive, and intensity. It will be interesting to see how top-dog Cymer adjusts its own gameplan and reacts to underdog Gigaphoton's push for bragging rights and increased legitimacy on the laser light source ballfield. Think AMD vs. Intel, Red Sox vs. Yankees, UMC vs. TSMC, and now Gigaphoton vs. Cymer.
|
Blogs 
