Intel Corp (INTC)

[Elmer Phud]

Semi -

What do you think of this?

http://www.eetimes.com/semi/news/OEG20030522S0063

Intel drops 157 litho from roadmap

By David Lammers

EE Times
May 23, 2003 (12:48 a.m. ET)

AUSTIN, Texas — Intel Corp. dropped a bomb on the lithography industry, informing its suppliers that it has removed the 157-nanometer lithography generation from its road map.
Peter Silverman, director of lithography capital equipment, said Intel will extend the 193-nm tools for the 45-nm node coming into production in 2007, where 157-nm tools were expected to play a role.

Technical challenges, including the quality of the calcium fluoride material required for the lenses, the lack of a feasible pellicle, and challenges with the 157-nm resists, all led to Intel's decision. It means that 193-nm tools will be used for the critical layers of the 90-, 65- and 45-nm generations at Intel.

After that, extreme-ultraviolet (EUV) lithography will be the "preferred candidate" for the 32-nm node, Silverman said, adding that Intel is keeping its options open there as well.

Intel's decision was based on "technical, not economic considerations," Silverman said, primarily the availability of commercial quantities of the calcium-fluoride material needed to make the lenses for the 157-nm scanners. Also, "the 157-nm resists are a long ways away from being production-worthy," he said.

Though Intel was one of the early backers of the 157-nm development program, Silverman said there is now a "complete consensus" within Intel that the 157-nm technical issues will not be solved in time for the 45-nm node. Intel can extend 193-nm tools with optical-proximity-correction mask techniques and lenses with a high numerical aperture.

The 45-nm node moves to volume production in 2007, but development is underway now. The half-pitch of the so-called 45-nm node actually is closer to 70 or 75 nanometers, he noted.

Silverman said Intel does not plan to use immersion techniques to extend 193-nm lithography. International Sematech has been coordinating an effort to extend optical lithography by immersing the lenses in liquids, which serves to enhance the resolution considerably.

"Immersion 193 is at best a one-generation tool that would barely cover the needs of the 32-nm node. The advantage of EUV is that it is an extensible technology that could be used for the 32-nm node and several generations beyond that," he said.

"This is a major bombshell for the lithography industry," said a source at a major lithography vendor, who asked not to be identified. An immersion lithography workshop scheduled for July could result in vendors endorsing "wet" 193-nm tools as a feasible technology, he said.

Doug Marsh, a vice president for investor relations at ASML Lithography, told the Deutsche Bank semiconductor conference in mid-May that yields on the calcium-fluoride production lines were in the 1 percent range. That remark buttresses Intel's decision. If ASML -- now the leading lithography vendor -- were to withdraw support for the 157-nm tools, coupled with Intel's decision, it could spell the death knell for the 157-nm effort.

For post-optical lithography, ASML is investing heavily in EUV lithography, while Nikon Corp. is putting much of its effort into electron-beam projection lithography.

Ken Rygler, a consultant, said Intel's decision means that it will need to push hard on the lithography "k" factor. Phase shift masks are expensive, and pushing the numerical aperture of the lenses incurs the penalty of reduced depth of field, which is need for vias and contact holes, Rygler noted.