AMD’s 2020 Vision for 450mm: Leave inefficiency behind

by Pete Singer, editor-in-chief

July 22, 2008 — There is a lot that can still be done to improve the efficiency of today’s 300mm factories. An early move to 450mm would only carry those inefficiencies forward another generation. That was the main message in an interview with Gerald Goff, principle member of technical staff for Advanced Micro Devices (AMD) in Austin, TX, during last week’s SEMICON West.

Goff described the company’s experience with the transition from 200 to 300mm, which was done only the keep up with leading edge process technology. “It’s not our perspective either that 450mm is never going to happen. It seems like it’s going to happen inevitably but the bottom line is that we just think the 2012 vision that’s out there is premature,” Goff said. “We don’t put a real date on it but we like to say we have 20-20 vision, but it’s 2020 and beyond is the timeframe when we will actually start to see some real life to it.”

Goff said “if you look at it from a holistic perspective, there’s really no point in taking all the inefficiencies we have today in our factories with us to 450. There are a ton of things we can be doing to eliminate inefficiencies in our factories at 300mm that are directly applicable to a 450mm fab. If we do the exact same thing that we’re doing today it’s going to multiply those inefficiencies by the square centimeters of the silicon we produce.”

Although it was not called 200mm Prime at the time, Goff said AMD’s transition to 300mm was not driven by a desire to improve fab efficiency, but only by the need to keep up with the latest technology that was only available on 300mm tools. “In Fab 30, we had a very productive 200mm fab. It was your conventional early 200mm fab which was open cassette, sub-Class 1 complete fab environment. We were doing fine, we were getting the volumes that we needed, we had the efficiencies. We had the lowest cycle time in the industry per Sematech data per mask layer. We were doing really well in that space and we had the potential to make some good money out of it. The problem in the 300mm space was all the R&D money shifted from the chambers at 200mm to the chambers at 300mm so the technology ran out. We could have still been very productive in a 200mm factory. We just couldn’t do what we wanted to do because we were leading edge technology wise, so we were forced to be shifted into the 300mm space.”

Goff said he sees the same thing happening moving from 300 to 450mm. “There are only so many R&D dollars that are available from either our side of the fence or the suppliers side of the fence.” Those R&D dollars are going to be put on where the industry decides they are most needed. “If we go focus on a wafer size change and forget about things like EUV and other process technologies that we’ve got to still work on 300mm — not to mention all the efficiency gains and white space eliminate that we can go through as far as we operate our factories then we’re going to be quickly forced into a 450mm environment if the R&D money starts going that way. This doesn’t really seem to make sense to us,” he said. — P.S.


Easily post a comment below using your Linkedin, Twitter, Google or Facebook account. Comments won't automatically be posted to your social media accounts unless you select to share.