In a natural way, several of Cymer's customers are reacting to the company's more aggressive pricing stance on new leading-edge laser systems by working with Gigaphoton to enhance its second-source position in an effort to generate some pushback leverage on Cymer (i.e., channel power). The extent to which this works will depend on Gigaphoton's ability to compete at the leading edge with Cymer, which we doubt it can sustain. It hasn’t proven its ability to do so in over the past 10 years. All in, this is exactly what happened when 193-nm ArF systems were in the evaluation phase. Several chipmakers used low-cost single-chamber Gigaphoton lasers for evaluation as the power and throughput requirements were limited. However, once volume chip production began, Cymer was dominant.

The historical competitive situation between Gigaphoton and Cymer is cyclical and accordionlike. Specifically, Cymer innovates and Gigaphoton follows. As Cymer’s innovations ramp into volume production, the company enjoys a period of relatively unfettered market dominance at the leading edge. Over a period of a few years, Gigaphoton begins to catch up and closes the market-share gap a little. As such, Cymer’s total chipmaker installed market share in dollars has historically been 80% to 90%, with the past few years in the 83% range as Gigaphoton began to experience some market success with its dual-chamber MOPO 193-nm ArF technology versus Cymer's innovative MOPA dual-chamber 193-nm ArF technology.

As 248-nm KrF and dry 193-nm ArF become larger trailing-edge markets, it is distinctly possible that Cymer’s “normalized” market share could trend down into the 70-80% range. However, the mix shift to immersion and ultimately EUV-based laser light sources suggests substantially above wafer fab equipment industry average growth for lithography light sources on major technological innovation, and therefore ASP expansion. It should also occur with limited competition (if at all for Cymer), at least initially, and solid above-corporate-average margins (i.e., 50%+) we think....

...Will Cymer's ring technology be as important is [sic] its innovation earlier this decade to the dual-chamber 193nm ArF MOPA approach, which caused Gigaphoton to go back to the drawing board as it was pushing a single-chamber approach? As a result, Gigaphoton was pushed back a few years. It is now catching up on the dual-chamber technology and is getting some incremental penetrations, including immersion, mainly for evaluation systems. We think a similar scenario is likely to unfold again....

...As we stated above, when Cymer introduced its dual-chamber MOPA technology Gigaphoton had to execute a total system redesign that took several years to ultimately commercialize in decent volume. We expect a similar scenario, assuming Cymer’s ring technology becomes the de facto industry standard later this year and into 2008. If this occurs, it could take several years to complete as Cymer believes that Gigaphoton will have to do a major system redesign to push the ring technology into its MOPO chamber approach.