While attending an ASML presentation at SEMICON Korea, I noticed a tidbit on one of the slides. A key part of their future EUV roadmap has been growing the power source to 800 watts or more. But it had not been super clear to me how they were going to do that, other than maybe raising the speed of the tin pellets. Then I saw the slide and it showed something new, at least to me. Three laser pulses rather than two. Now how about that? No, this is not an Onion article. This looks to be real. In this video, I want to talk about lasers and EUV light sources. And how three is better than two.
Conclusion
So what does it mean? Achieving a three-pulse structure will potentially require multiple laser sources, precise time control down to the nanosecond level, and high control/optimization of the pulses' parameters. If ASML manages to pull this off, then it implies as much as a 30-50% jump in conversion efficiency, as well as significant improvements in debris generation. And now suddenly, 800 or even 1000 watts looks within reach. This is a big deal. For years, EUV machines were starved for EUV light photons, forcing us to use it only where necessary. Now there is a possible future ahead where that starvation turns into an abundance. The possibilities that holds for semiconductor manufacturing are exciting. Assuming they can do it, of course.
Conclusion
So what does it mean? Achieving a three-pulse structure will potentially require multiple laser sources, precise time control down to the nanosecond level, and high control/optimization of the pulses' parameters. If ASML manages to pull this off, then it implies as much as a 30-50% jump in conversion efficiency, as well as significant improvements in debris generation. And now suddenly, 800 or even 1000 watts looks within reach. This is a big deal. For years, EUV machines were starved for EUV light photons, forcing us to use it only where necessary. Now there is a possible future ahead where that starvation turns into an abundance. The possibilities that holds for semiconductor manufacturing are exciting. Assuming they can do it, of course.
