It's 2012: Time to get serious about 450mm?
Research Vice President,
The effort to actively engage equipment companies in the serious development of tools for 450mm wafers got a major boost last fall with the formal announcement of the investment of serious money by four major semiconductor manufacturers and the state of New York aimed at 450mm development at the Albany Nanotech facility in upstate New York. The $4.4B effort, which includes R&D work, will be funded by contributions from Intel, Samsung, TSMC, Globalfoundries, IBM and the State of New York.
Prior to the announcement of this new facility, 450mm development efforts had been bumping along at a slow pace that might see a production ready fab by the end of the next decade. The new facility indicates that the major semiconductor companies who had been the driving force for 450mm development realized that they would have to do something to kick the pace of development into high gear if they were going to see a production fab before 2020. The location of the new facility is significant. Albany NanoTech has already been well established as a successful cooperative R&D consortium where equipment companies and semiconductor manufacturers are able to work together to develop advanced processes while maintaining the necessary protections of intellectual property. Building on the existing levels of cooperation has the potential to dramatically reduce the costs of 450mm development to the point where a reasonable payback of R&D investment is possible.
But what this facility represents is a very necessary step in the initial concept and feasibility studies to determine if 450mm has a chance of achieving its very ambitious cost goals. It does not guarantee that these goals will be met, or that the very real engineering challenges can be overcome to the degree necessary for implementation in volume production. And the challenges are significant.
The biggest challenge is whether lithography can meet the cost targets necessary to achieve the stated goal of a 25-30% reduction in the cost to manufacture silicon compared to 300mm. So far ASML has not made a commitment to development 450mm tools ??? they are currently focusing all their efforts onto getting EUV production worthy on 300mm. To further complicate matters, 450mm production will require both EUV and 193 immersion tools, with significant engineering resources needed for each.
Gartner projects that if everything works right, the first production 450mm fabs will appear around 2019 ??? as the industry is moving to the sub 10nm feature size for leading edge production. This raises an interesting question: with the challenges of implementing sub 10nm production, will semiconductor manufacturers be willing to risk it all on untried 450mm tools, or will they demand parallel process developments on 300mm to hedge their bets? In addition, how long will conventional silicon processes developed for sub 10nm production continue before silicon hits its limit? Will this be sufficient to guarantee a reasonable rate of return on equipment R&D investments?
And finally there is the question of exactly how many 450mm fabs will be built? Already only a handful of semiconductor companies are building new fabs, and the list will likely get smaller, not larger. Will this be sufficient to justify 450mm development costs?
There are still lots of questions which need to be answered and engineering problems which need to be solved, but the increased collaborative efforts at Albany NanoTech indicate that the industry is getting serious about trying to find answers.
Solid State Technology, Volume 55, Issue 2, March 2012