Most foundries have yet to start buying the capital equipment needed for the 14/16 nm node, which for many will be the first to support FinFETs, says Trafas of KLA-Tencor. Gear companies hope the orders start coming in the fall.
Meanwhile most fabless designers are testing out the waters in 20 nm to get their first taste of the new multi-patterning techniques.
"Some companies have made the transition to 20 nm, and others decided to skip it and go to 16/14 nm," Trafas says. "While there is 20 nm spending happening, I think 16/14 nm -- due in part to higher complexity -- will be a higher investment level."
Indeed, he says, one of the big questions many capital equipment execs will bring to this year's Semicon West event on July 7 is, "When will the 16/14 nm investments begin?"
Work on that node "is very early, focusing on initial test devices such as SRAMs. A unique issue is at that node the back-end design rules are getting tighter with multi-patterning challenges at metal-1 and above."
Most vendors looking ahead to 10 nm expect not to have extreme ultraviolet lithography available, says Trafas. "They want to implement EUV at 10 nm if it's available maybe on a single layer or two, but most have efforts to do 10 nm with or without EUV."
For its part, KLA is researching how to improve all its products' subsystems for the tighter critical dimensions and overlay measurements needed for 10 nm.
Applied's Bencher says engineers see ways to use today's multi-patterning techniques to create 10 nm chips. But beyond that, "7 nm looks completely brutal. You either must have EUV or some architecture change like multi-die packaging so the system scales."
Next page: The need for 3D design