of failed litho
expect R. Fabian Pease to mince his words. The Stanford University professor
was in fine fettle as a plenary session speaker at this year's SPIE
Microlithography event in San Jose. His talk, "Lithographic
Technologies That Haven't (Yet) Made It—Lessons Learned,"
had the room rippling with laughter and bobbing with knowing nods.
showed one slide in which he tried to pick a failed litho technology
to fit every letter of the alphabet. The dozens listed included a handful
crossed off or dotted-lined to indicate that they had "made it"
or "probably will work." Not one to hide his own missteps,
the good professor also showed those failed technologies that he had
can be learned?" he asked. "To displace the old, the new must
be 10 times better. What
do we mean by 'better'?" He explained that "better" must
combine the ability to print the smallest features, with the best controlled
feature size, best depth of focus, fastest throughput, and best overlay,
and it must be the least costly in terms of capital and operating expenses
per assured good pellicle. Pease sees nanoimprint as the only foreseeable
replacement for optical.
one barbed comment, Pease characterized Intel's sacred next-gen litho
cow, EUVL, as "soft x-ray technology," calling it "the
most egregious example of a wolf in sheep's clothing." A believer
in using "wavefront engineering" to extend optical litho,
he offered a "dangerous extrapolation" showing that the entire
Intel production run of processors would be required for the needed
computing power to do resolution-enhancement techniques in 2030—"just
to put the resist edges in the right places!"