Although Globalfoundries competes head to head with TSMC, UMC and other dedicated foundries, Jelinek maintains that the company's primary competition is Intel. Globalfoundries is now No. 2 in foundry sales, but the company still trails TSMC by a wide margin. Globalfoundries is seeking to establish itself as the process technology leader among foundries largely because its largest customer, Advanced Micro Devices Inc., is Intel's chief competitor in the market for PC processors.
"They are not going to try to grow and become No. 1 foundry player in
the world. They know they cannot get there," Jelinek said. "What they
can do is become the technology leader in the space, and that's what
they are trying."
According to G. Dan Hutcheson, chairman of market research and consulting firm VLSI Research Inc., no chip maker has ever moved to a new process technology node two years in a row. "if they can pull this off, it will be a tremendous achievement," Hutcheson said.
Subramani Kengeri, head of the advanced technology architecture for Globalfoundries' office of the CTO, said the similarities between the company's 20-nm low power process and the 14-nm XM process would enable customers to leverage much of the design work for the 20-nm process on the new process. About 7,000 design rules are the same between both processes, while about 60 new design rules will be added to account for the FinFET, Kengeri said.
The 14-nm XM process utilizes the same middle and back end of line processes as Globalfoundries' 20-nm low power process, which is set to go into production next year.
Technology development of
the 14-nm XM process is already underway, with test silicon running
through Globalfoundries' Fab 8 in upstate New York., the company said. Early
process design kits are available now, with customer product tape-outs
expected in 2013.
You're a little more optimistic than me. I think they're overconfident given that 14nm is the first node they'll use FinFETs at, along with the fact they haven't begun any kind of volume with 20nm (even 28nm is fresh out the door).
But we'll see. If Globalfoundries pulls this off they'll be in a strong position.
I think you have to give GF credit for boldness. As the comments above allude to, they had their share of yield issues at 32-nm. The GF executives I spoke with said they wanted to wait until the process design kit was available to make the announcement. To G. Dan Hutcheson of VLSI Research, the fact that the process design kit is available is the key to the announcement.