GlobalFoundries has confirmed its 14nm process as the one that will see the foundry introduced finFET-based transistors, claiming that its approach is optimized for mobile devices, in contrast to Intel which is attempting to hit sweetspots in both desktop/server and mobile systems.
The foundry is aiming for a 40-60 per cent improvement in battery life compared to designs that use planar transistors at the 20nm node. To some extent the 14nm process will be a shrink of the 20nm technology with the addition of the finFET devices.
The company claimed technology development is already underway, with test silicon running through Fab 8 in Saratoga County, New York. Early process design kits (PDKs) are available now, with customer tape-outs expected in 2013.
“We have more than a decade of finFET R&D to build on as we prepare to bring this technology to production,” said Gregg Bartlett, CTO at GlobalFoundries.
According to the foundry the aim is to allow companies to port over existing IP in a relatively straightforward manner. This echoes what Intel did with its first finFET designs which made heavy use of process-migration tools.
One issue with finFETs is that they do not readily support multiple threshold-voltage options because of their, normally, undoped nature. At the VLSI Technology Symposium, IBM and GlobalFoundries collaborated on a paper that hinted channel doping could be used to increase the flexibility of finFETs for use in SoCs. It at least seemed to work at 20nm – it will be interesting to see if this work carries through to the production 14nm process at GlobalFoundries.