Issue



Euro Focus


05/01/1999







Siemens' 300-mm plans

Declaring that "the future starts in 2001," Siemens VP of world DRAM manufacturing outlined a tentative roadmap for 300-mm production that would call for a volume production facility to be ready for equipment by mid-2001, and for all of the company's DRAM production to be on 300-mm wafers before 2004.

Speaking at the Semi-sponsored ISS-Europe conference in Rome, Harald Eggers said current plans call for Siemens' first volume 300-mm fab to have a 5000 wafer/week capacity. No site has been chosen. The facility would produce 1-Gbit DRAMs, starting with 140-nm or 130-nm design rules and a processing cost per unit area 30% below an equivalent 200-mm facility. The 300-mm wafers could hold 338 of the 185-mm2 devices, compared to 140 on a 200-mm wafer.

Eggers did, however, make the point that "these are plans," rather than firm commitments. Productivity goals must be met at the Semiconductor300 (SC300) pilot line in Dresden (a joint venture with Motorola), and prices on tools must come down.

Siemens hopes to achieve sustainable yields of 50% or more on 64-Mbit devices at SC300 by June 2000. The knowledge gained at SC300 will be combined with research from a separate Fab of the Future (FoF) project as the first 300-mm production line is planned.

Noting that the first lot of wafers run through SC300 yielded good parts, Eggers commented, "300-mm equipment is not only available, but can run with suitable cycle times." Of a group of 58 tools at SC300, 13 were categorized as "early production," 23 as "beta," 13 as "alpha" and nine as "prototype." Wet process equipment was generally the most mature group, followed by etch, lithography, and thin film.

Meanwhile, the FoF project is working on creating and implementing "a new concept of productivity, manufacturing, logistics, CIM/automation, layout, design, and construction," Eggers said. This concept is to be fully implemented in Siemens' second 300-mm production fab, and later propagated to back-end packaging, assembly, and test facilities. One possible innovation would be the use of space above the cleanroom ceiling for storage of work in progress, eliminating the need for WIP stockers in cleanroom areas.

FoF research and projections have found that if a 200-mm wafer using 200-nm design rules costs 1000 arbitrary units to process, an equivalent 300-mm wafer should cost 1610 units, while one carrying 130-nm features would cost 1820. This would result in significant savings per area processed, since the 300-mm substrate is some 2.3 times larger.

In order to meet these goals for the second production fab, however, a 17% gap between target equipment prices and current price estimates will have to be closed. Currently, the biggest problem area is ion implant, which is running about 50% over target cost. Wafer test is about 42% over, PVD 35%, CMP 23%, and metrology 22%. On a brighter note, etch and CVD are at target price, and lithography (which accounts for 30% of the overall budget) is just 10% over. Improvement is anticipated through process simplification and "reduction of estimated equipment prices," said the presentation. "We will talk to our suppliers, ask about the productivity gap, and say they have to meet expectations or it won't work," said Eggers. Costs on fab infrastructure, such as cleanrooms and buildings, show an overall 21% gap between target and estimated price. "This is another task for the FoF project," he said.

Eggers wrapped up by saying, "There is a realistic chance for a high-volume fab to meet the productivity targets and close the 17% gap ... but the cost model leaves no room for nice accessories."

While the Siemens plan must be classified as tentative, it does mark the most detailed 300-mm roadmap to be publicly disclosed, and reinforces a sense in the equipment industry that Siemens will be an early adopter of the larger wafers. As Siemens Microelectronics prepares to be spun out from its parent and focuses its efforts on DRAMs, an aggressive move to 300-mm could provide it with a competitive advantage over Micron and Asian rivals. - P.D.

Steag acquires AG Assoc.

Steag Electronic Systems GmbH, Essen, Germany is acquiring RTP systems maker AG Associates, San Jose, CA, in a stock tender deal worth an estimated $42 million. The privately held Steag made a cash tender offer for all AG Associates' outstanding shares, paying $5.50 in cash for each share. There are about 6.2 million shares of AG's stock outstanding. Kirk Johnson, AG CFO, said the deal is roughly valued at between $42 million and $43 million, after taking into consideration Steag's assumption of AG's bank debt.

AG will become a private, wholly owned subsidiary of Steag, combined with Steag's existing RTP operation in Germany, Steag AST Elektronik. Johnson said both RTP operations will continue to operate in their respective locations, at least initially; details of Steag's long-term goals for the RTP business have not yet been finalized. "Most of the work (to date) has gone into getting an agreement signed."

The two firms have been exploring a possible tie-up since at least November, when they disclosed that discussions were ongoing. "When we sat down with Steag and looked at our respective companies, we found that we had very little customer overlap and shared a common vision," said AG CEO and chairman Arnon Gat. "Looking forward, we knew that in order to remain competitive, AG would have to find the right partner." The two firms have also not determined whether Gat and AG's current management team will remain with AG under Steag, Johnson said. - C.L.

While the Siemens plan must be classified
as tentative, it does mark
the most detailed 300-mm roadmap to be publicly
disclosed, and reinforces
a sense in the equipment industry that Siemens will
be an early adopter
of the larger wafers.