Dongbu's DL013GD 0.13-micron process will employ up to eight layers of metal and poly transistor gate lengths as small as 0.08 micron, said Ed Chow, vice president of technology support for the Korean foundry company. The process will offer FSG (fluorine-doped silicate glass) and low-k inter-metal dielectric options, and the company expects to offer both "foundry-compatible" and a version that's based on technology from Toshiba Corp.--Dongbu's partner.

Dongbu's update on its 0.13-micron plans comes at a time when a number of major silicon foundry companies are reportedly having trouble reaching acceptable yields on their copper-based 130-nm (0.13-micron) processes. In nearly all cases, Asian foundry companies have not admitted to any problems publicly, but responses in recent conference calls indicate that 0.13-micron technologies are still somewhat in "debug," or "early-learning/yield-improvement" stages.

Moreover, reports of widespread yield problems at 0.13-micron are raising some questions about the ability of third-party foundries to quickly deliver next-generation processes as more IC suppliers opt to shut down their own wafer fabs for contract manufacturing services. Part of the problem could be related to design issues as well as production steps, according to industry experts.

For example, sources believe that Taiwan Semiconductor Manufacturing Co. Ltd. (TSMC) is having some problems getting its 0.13-micron technology up and running after rolling out the process last year. Some chip makers complain of lower-than-expected yields with TSMC's 0.13-micron process, while still others seem satisfied with the technology, according to sources in the industry.

At present, TSMC offers two separate versions of 0.13-micron technology. The first version is a copper-based technology, with an FSG (fluorine-doped silicate glass) option. Meanwhile, the second version is a copper-based technology, with a low-k dielectrics option.

TSMC is still having some "reliability" issues with the 0.13-micron process, based on low-k dielectrics, confirmed Shang-yi Chiang, senior vice president of research and development for the Hsinchu-based foundry giant, in an interview with SBN last week. "We hope to solve that within the next three months," Chiang said.

The copper-based technology with FSG is not experiencing any "reliability" problems, he insisted. This process is currently "being qualified by customers," he insisted.

While TSMC is ironing out the glitches with its 0.13-micron process, the company last week disclosed plans to put its 0.10-micron technology into production by the third quarter (see Jan. 30 story).

But TSMC is not the only foundry having some problems with 0.13-micron technology. For example, last week one analyst asked Advanced Micro Devices Inc. if it was concerned about reports of problems in 0.13-micron copper processes at its new processor foundry partner, United Microelectronics Corp. in Taiwan. At first, AMD chairman and CEO W.J. (Jerry) Sanders responded back to the analysts, saying, "First of all I don't know who is having a problem with 0.13-micron."

But then Sanders expressed full confidence in UMC's ability to develop and produce leading-edge processes in foundry fabs. "AMD has been saying for 33 years, 'people first and products and profits will follow.' We could modify that and say, 'people first and processes and profits will follow,'" said the chief executive officer. "It turns out there is a great team of people at UMC," added Sanders, referring to the world's second largest silicon foundry company, which will be producing PC processors for AMD later this year. UMC and AMD are also planning a joint 300-mm fab for 0.065-micron processes in Singapore by 2005 (see Jan. 31 story).

"We think we make a great team," Sanders said. "They do the basic platform process and we are the leaders in high technology and high-performance transistor development."

UMC officials also responded to questions about 0.13-micon yields, but they did not deny that the foundry company was still in an early stage of tweaking the technology. "With the 0.13-micron technology, we are in the early production phase, and we actually already have a lot of customers shipping samples or the products to their customers now," said Fu Tai Liou, chief officer of sales and marketing.

"We are still in the early stage, but we are in pretty good shape for this 'pre-production' or 'early production' stage right now," added the sales and marketing chief who was previously UMC's chief technology officer.

Questions about 0.13-micron yields also surfaced during last week's quarterly-results conference call hosted by Chartered Semiconductor Manufacturing Pte. Ltd. in Singapore. The world's third largest pure-play silicon foundry company announced last summer that it had fabricated its first 0.13-micron copper IC for a customer (see July 16 story), but nearly nine months later it still has not officially sold a 0.13-micron wafer for revenue.

"We are literally days away from making our first revenue shipments in 0.13-micron, I believe," said Barry Waite, president and CEO of Chartered, during a conference on Jan. 29. "I'm expecting 'full qual' in the late third quarter time period. We will be generating and ramping revenues in late third quarter going into fourth quarter, and that is on a full copper 0.13 micron process."

Chartered officials indicated that the process continues to improve in yields as its initial wafers are fabricated in 8-inch fabs, but no details were provided during last week's conference call. The current copper process uses FSG for dielectrics in the interconnect, and the company continues to work on alternative lower-k approaches, Waite said, without providing details. Chartered is teamed with Agere Systems Inc. in the development.

The difficulties of ramping 0.13-micron continues to keep the door open for other startup foundries, which got into the business using 0.25-micron technologies in the past couple of years. The quarter-micron segment has been hit hard by price erosion in the current downturn, causing foundries to push harder for 0.18- and now 0.13-micron business.

Dongbu believes its low-voltage 0.18-micron process (DL018LVT) is hitting the market at the right now. The company said the process has demonstrated "high yields" after 1,000 hours of burn-in.

"Our recent DL018LVT process launch put us solidly in the middle of the current 0.18-micron CMOS design window," declared Peter Hillen, executive vice president of worldwide business development for Dongbu in Mountain View, Calif. "In a similar fashion, we expect our DL013GD 0.13-micron process to launch at an ideal time later this year to meet the growing production demand for next-generation system-on-chip (SOC) designs for wireless, networking and advanced consumer electronics applications."

The Korean foundry also expects to provide early adopter access to both digital and mixed-signal technologies. "With a 1.2-V core and I/Os of 2.5-V and 3.3-V, plus a 6-T SRAM cell measuring only 2.40 square microns -- our 0.13-micron process will be ideal for implementing SOC designs that integrate high density SRAMs," stated Dongbu's Chow.

--J. Robert Lineback reporting from the U.S.