SAN FRANCISCO The 450-mm debate rages on.
Intel, Samsung and TSMC are pushing for 450-mm fabs by 2012 or so. On Thursday (July 17), the fab-tool community will hold a panel discussion on 450-mm technology. At the Semicon West trade show here, SEMI--and many fab-tool vendors--earlier this week telegraphed their position on 450-mm, reiterating that the economics do not add up for the next-generation wafer size.
Representing a large share of the fab-tool community, SEMI also released its long-awaited white paper on 450-mm technology. In the paper, the trade group said it could cost ''well over $25 billion" in R&D to fund 450-mm technology.
"The semiconductor industry cannot afford to make such an investment based on future 'expectations' without an objective analysis of cost and benefit,'' according to the trade group.
Based on its simulations, SEMI's Equipment Productivity Working Group (EPWG) concluded that ''450-mm wafer scale-up represents a low-return, high-risk investment opportunity for the entire semiconductor ecosystem; 450-mm should, therefore, be an extremely low priority area for industry investment.''
Stanley Myers, president and CEO of SEMI, said 450-mm technology is somewhat like the industry's ill-fated efforts in 157-nm lithography. The IC industry spent billions of dollars in 157-nm development, but Intel Corp. dropped its program and the rest of the industry followed suit.
In other words, the IC industry led the equipment community down a ''blind alley,'' Myers said. Like 157-nm lithography, the prime backers behind 450-mm technology are also leading the industry down a ''blind alley.''
Leading fab-tool makers agree, saying it is too expensive. Besides, there are still productivity gains to be made with current 300-mm fabs, said Tom St. Dennis, senior vice president and general manager of the Silicon Systems Group at Applied Materials Inc.
"We are not investing in it today,'' St. Dennis said at an event, adding that Applied still does not ''see the benefit.''
On the opposite side of the fence, chip-making consortium International Sematech recently provided an update on its next-generation 300- and 450-mm programs, saying that they are on track and making steady progress.
Some tool vendors said that they would devise 450-mm tools--if a chip maker is willing to pay for the R&D. But at the SEMI press event, fab-tool vendors spoke out against 450-mm.
John Ellis, vice president of global standards and technology for SEMI, said Sematech's figures simply don't add up. But, he said at the event, "we aren't here to beat up on ISMI.''
Because Intel, Samsung and TSMC are pushing for 450-mm, fab-tool vendors are not rushing to develop 450-mm tools. Vendors will continue to advance their 300-mm tools. "We don't see reason to change our priorities,'' said Skip Miller, director of strategic marketing at ASML Holding NV.
Besides, going to 450-mm provides no real benefit for the industry, said Masayuki Tomoyasu, senior vice president and chief engineer at Tokyo Electron Ltd. (TEL)
''Cost and risk are too great'' for 450-mm, said Walter Class, a consultant, who represents Axcelis Technologies Inc., at the event.
''Scaling of plasma tools is non-trivial'' when going to 450-mm, said Dave Hemker, vice president of new product development at Lam Research Corp., referring to CVD, etch and related technologies.
Hemker said that he is more worried about developing 300-mm tools for the 32-nm node and beyond. 450-mm is only geared for limited customers: TSMC, Intel and Samsung. Vendors like Lam can't afford to build tools for a ''subset'' of customers, he said.
So, when does it make sense to build 450-mm fabs? Perhaps in the distant future. But there is no reason to have a 450-mm fab for the foreseeable future, said Iddo Hadar, chief technology officer for the Foundation Engineering Group at Applied Materials Inc.
"We're not saying never. It doesn't make sense now," Hadar said.
SEMI's white paper can be found here.