2019-2022 EUV wafer output

[Fred Chen]

From SPIE paper 1249406: C. Smeets et al., 0.33 NA EUV systems for high volume manufacturing

 

[Fred Chen]

Probably 4 possibilities:
1. Lots of idle time (e.g., from lower demand)
2. Lots of downtime (probably less likely, >90% uptime reported)
3. High dose (100 mJ/cm2 or more), which would require a suitable resist process, but also would have more secondary electron dose (blur) as well
4. @Tanj has brought up an interesting interpretation of multiple layer exposures per wafer. 40 wph could be 120 wph averaged over three layers, for example. Not quite the expected number of layers, for sure. Or even a single layer with multiple (e.g., 3) exposures, still odd.

We also haven't considered R&D wafers, which might also be substantial, e.g., resist characterization.

 

[Fred Chen]

The 5 wafer output data points can be fit with a quartic polynomial fit. When that is done, we can better pin the quarterly wpd/tool.

 

[jms_embedded]

Could you cite any other comparable information for lithography machines? (wafers per hour or wafers per day for one layer)

 

[Fred Chen]

Could you cite any other comparable information for lithography machines? (wafers per hour or wafers per day for one layer)

Sure, more on wafers per hour here: Assessing EUV Wafer Output: 2019-2022