SI CHINA     SI JAPAN
Login  |  Register          Free Newsletter Subscription
Subscribe
Email
Print
Reprint
Learn RSS

Real Men/Women Do Have Fabs

Laura Peters, Editor-in-Chief -- Semiconductor International, 4/30/2008 7:54:00 AM

Following on the now-famous line from Jerry Sanders, “Real men have fabs,” Tom Sonderman, vice president of manufacturing technology at Advanced Micro Devices (AMD, Sunnyvale, Calif.), declared at yesterday’s SEMI Strategic Business Conference (Napa Valley, Calif.) that “real men and women do have fabs.” However, AMD is increasingly relying on partners, including IBM (East Fishkill, N.Y.) and Chartered Semiconductor Manufacturing (Singapore). In East Fishkill, IBM is providing the R&D to bring the baseline technologies directly into production so that AMD does not have to run a separate pilot line. This collaborative program for device integration and process qualification allows low-defect densities to be immediately achieved, so faster yield ramps of its 45 nm technology in the first half of 2008 will be possible (Fig. 1).

1. By eliminating the pilot line, higher yields are achieved sooner. (Source: T. Sonderman, AMD)

Sonderman commented on some specific process technologies. “Double patterning is counterintuitive to lean thinking. AMD is reportedly the first microprocessor vendor to ramp immersion lithography in a fab, which is occurring in Dresden. AMD does have what it is calling a 'gate-first' high-k/metal gate option for 45 nm, but it will be offered on the second generation of 45 nm technology. We were able to achieve the performance targets without implementing high-k/metal gate, so we will introduce it when it is appropriate,” Sonderman said. He noted that, in the fab, AMD makes continuous improvements to the transistor.

AMD heavily invests in all aspects of lean manufacturing, including new scheduling programs, one of which will be implemented in New York in the near future. From February 2005 to November 2007, the Dresden fabs were able to dramatically increase wafer starts and labor productivity and reduce cycle time and processed wafer cost (Fig. 2). “Even with the results we’re getting, we think we’re only scratching the surface,” Sonderman said. He talked about the Next Generation Factory (NGF), also called 300mmPrime, and some of the targets for that program. “A lot can be gained by optimizing batch size and better scheduling. Equipment should be more flexible and predictable.” He said the “nirvana” of cycle time is 1 day per mask layer, but with optimized batch sizes, scheduling and next-generation material transport, AMD believes <0.7 day per mask layer is possible.

2. Dramatic improvements in efficiency and productivity through aggressive and consistent process improvement using lean methodologies
Email
Print
Reprint
Learn RSS

Related Content

Related Content

 

By This Author

SPONSORED LINKS


SI Resource Center


 
Advertisement
SPONSORED LINKS

More Content

  • Blogs
  • Podcasts
  • Videos

Blogs

Podcasts

Videos

  • CMOS and Beyond: Surface Prep at Nanoscales
    Now Playing: CMOS and Beyond: Surface Prep at Nanoscales
    At SEMICON West 2006, Alex Braun interviews Ahmed Busnaina, director of the NSF Nanoscale Science and Engineering Center for High-Rate Nanomanufacturing at Northeastern University. Busnaina gives his perspective on the longevity of CMOS, nanotechnology, and surface preparation at nanoscales.
    Video Table of Contents  More Videos >>
Advertisements





NEWSLETTERS
Plug in and get the latest SI news, trends and industry updates delivered free, directly to your inbox!

SI NewsBreak and Special Reports (Weekdays)
Wafer Processing Report (Monthly)
Lithography Report (Monthly)
Metrology Report (Monthly)
Clean Processing Report (Monthly)
Packaging Report (Twice Monthly)
©2008 Reed Business Information, a division of Reed Elsevier Inc. All rights reserved.
Use of this Web site is subject to its Terms of Use | Privacy Policy
Please visit these other Reed Business sites