SAN FRANCISCO - Amid a flurry of speculation, Intel Corp. confirmed that its new fab in the United States is being constructed for the 450-mm wafer era.
The recently-announced fab will be 450-mm compatible, but it could also support 300-mm tools, said Mark Bohr, Intel Senior Fellow and director of process architecture and integration at Intel. If the 450-mm tools are not ready, Intel could use 300-mm machines in the new fab, it was noted.
As reported, Intel will build a new R&D wafer fab in Hillsboro, Ore., and upgrade other existing U.S. facilities for 22-nm production at a total investment of between $6 billion and $8 billion.
The new development fab in Oregon, to be known as D1X, is slated for R&D startup in 2013. Speculation that Intel was plotting a new fab at its Ronler Acres campus in Hillsboro has been building for weeks. Some speculated that the facility would be a 300-mm or even 450-mm production fab. But Intel described the facility as just a ''development fab.''
One analyst recently decribedthe fab as ''450-mm ready,'' meaning that it will be constructed to support 450-mm tools-if or when the machines are ready.
Intel confirmed Tuesday (Dec. 7) that D1X is being readied for 450-mm. ''Intel is very interested in 450-mm,'' Bohr told EE Times. ''D1X is being (contructed) to be compatible with 450-mm.''
The fab tool community is also warming up to 450-mm. At one time, they did not want to devise 450-mm tools, because the cost is too high.
Now, there is a different story in the market. "I sense that some of the equipment vendors are interested in 450-mm,'' he said.
The 450-mm activity is heating up. ''I was surprised to see that that it’s just not the SIT consortia (Samsung, Intel, and Toshiba) that are now getting behind this technology. For the first time, I heard chip executives outside these three move from a position of never to 'not a question of if, but when,' '' said G. Dan Hutcheson, CEO of VLSI Research Inc., in a recent report. ''Right now, more than 90 percent of the equipment supply base is involved in some form of 450-mm development, though most these still hold public positions of never.''
Still, there’s a lot of work to be done in the 450-mm tool arena, he said. 450-mm fabs could appear in 2018-at the earliest, he added.
As expected its on now, and not the nikon is the supplier for the fabtool. Its kind of combination from applied, asml and nikon last. As nikon had serious problems with extreme litho its was the reason why intel initially announced different than expected.
No matter who says what, as the leader is in, every other big guy will follow him, so 450 is in place. However i dont think we could see 450 with sub-14nm as litho gets mad, and materials goes crazy. I wonder when 3D chip sets will be in place.
On collaboration to cut design and manufacturing costs certainly. On economics, Moore’s objective has presented an issue sub micron and all the time. Should industry double transistor density every 18 to 24 months on a pressed lithography schedule’s economics? Acknowledging Rock’s insight of doubling investment dollars each process generation has die economics become a fallacy? Where Intel has raced lithography inorganically how many nodes ahead to achieve its own process gains and too who’s benefit? When producing sequential short runs of surplus offer’s no economic profit too pay for this inefficiency over the long run? Too support producers or enablers of these subordinate economic effects? And now on trailing edge of CMOS average cost curve should development dollars be invested toward deflationary production scenarios on inflationary cost structures? Does 300mm+ acknowledge transition from industrial commercial art back towards applied science? Too bridge across a monopoly divide in development dollars toward all of our molecular futures? Meaning lithography, materials and etch but not harvesting on larger wafers. A monopoly aim whose time has past? Where competitors in this race have recognized their math error? Finally, IDM and end markets should pay more for this leading performance hardware enablement not less. Mike Bruzzone, Camp Marketing Consultancy
PS- I was being facetious above. It sounds like there has been more progress on 450-mm than I realized. I'm still not convinced that 450 will be in production anytime soon, or that it will ever have mass appeal, even for the handful of companies that can afford it. But time will tell.
more time passed between 300mm and whatevercomesnextmm, than between any two wafer size introductions in history. That's because it was so good at what it sought to do; reduce chip cost. Along the way, we also cut global warming gases per chip and made many other improvements.
Blog Doing Math in FPGAs Tom Burke 15 comments For a recent project, I explored doing "real" (that is, non-integer) math on a Spartan 3 FPGA. FPGAs, by their nature, do integer math. That is, there's no floating-point ...