Single page Print

More on the Fab 36 grand opening
I wrote up a brief news item about the grand opening event for AMD's new Fab 36, but I didn't have time then to say too much about it. Unfortunately, the time that I spent finding out about AMD's future plans overlapped with the various facility tours that the company was offering. As a result, I don't have much in the way of pictures or info from the fab itself. The grand opening's press conference and speeches did yield some interesting info, though.

Fab 36 is situated directly next to AMD's only other CPU manufacturing plant, Fab 30, in Dresden, Germany. Fab 30 manufactures 200mm wafers, while Fab 36 is capable of making wafers 300mm in diameter. 300 mm wafers have more than twice the surface area of 200 mm wafers, so they can hold quite a few more chips. Better economies of scale at 300mm make for much lower costs per chip than 200mm fabs, so Fab 36 should be a nice boost for AMD. CEO Hector Ruiz said that AMD can produce roughly 50 million microprocessors per year right now with Fab 30, and that AMD will be capable of producing over 100 million CPUs per year when Fab 36 reaches its full potential in 2008.

A fab employee shows off a 300mm wafer

German Chancellor Gerhard Schroeder pauses during his speech
to inwardly consider the prospects for invading France

AMD was eager to emphasize what it called Fab 30's "flawless" track record of execution, and executives stressed that Fab 30's success was a big reason for investing again in the same area. The company clearly wanted to take advantage of the team of skilled workers it had already assembled in Dresden. I'm sure the subsidies that the company got from state and federal governments—totaling roughly $500 million—didn't hurt, either.

Once Fab 36 is really churning out 300mm wafers, Fab 30 probably won't be making AMD's high-end microprocessors anymore. Ruiz said AMD was considering several options for Fab 30, including what he called "extension of the x86 footprint" (perhaps to mobile phones or hand-held devices?) or, interestingly enough, building chipsets—a business AMD hasn't pursued aggressively in recent years. He also mentioned the possibility of upgrading Fab 30 to newer technology at some point in the future.

The opening of Fab 36 is a major milestone for AMD, and we were pleased to come away from the event with a better sense of AMD's future plans for its CPUs. However, I'm still a little bit surprised that AMD hasn't disclosed a real roadmap populated by code names associated with various features and process technologies, as it has in the past. Obviously, AMD is playing its cards very close to its chest these days, for whatever reason—perhaps the transition to a new CTO, a desire to emphasize its present performance leadership rather than future products, or the need to keep more information confidential while pursing its various legal actions against Intel. At least now we know that AMD does have some specific plans that extend beyond the end of 2005, and we have some sense of how its future processors may look. 

Like what we're doing? Pay what you want to support TR and get nifty extra features.
Top contributors
1. Ryszard - $503 2. punkUser - $502 3. the - $306
4. SomeOtherGeek - $300 5. Ryu Connor - $250 6. doubtful500 - $200
7. Anonymous Gerbil - $150 8. cygnus1 - $126 9. danny e. - $125
10. SecretSquirrel - $125
Intel's Xeon D brings Broadwell to cloud, web servicesA big compute node in a small package 40
AMD previews Carrizo APU, offers insights into power savingsExcavator cores and other innovations to help improve efficiency 115
The TR Podcast 169.5 bonus edition: Q&A intensifiesYou ask, we attempt to answer 5
Samsung's Galaxy Note 4 with the Exynos 5433 processorA Korean import gives us a look at ARM's latest tech 110
Intel's Xeon E5-2687W v3 processor reviewedHaswell-EP brings the hammer down 114
AMD's FX-8370E processor reviewedEight threads at 95W 147
Intel's Core i7-5960X processor reviewedHaswell Extreme cranks up the core count 198
AMD spills beans on Seattle's architecture, reference serverCache networks and coprocessors 46