TSMC and Samsung, two competing foundries, have already successfully shipped 10nm wafers. They'll also be in the upcoming iPhone, which is a pretty high-volume part.
It's worth noting that node size refers to very different things between different manufacturers, we have very little idea about whats different/similar between TSMC/Samsung/Glofo/Intel's '10nm' nodes.
It's very possible that TSMC and Samsung are using a less technically challenging (and potentially less technically capable) definition of 10nm than Intel.
Or, it could be that TSMC and Samsung are actually technically ahead of Intel.
Yep, that's a fair point. In particular, because Intel isn't actually shipping anything at 10nm, all we have to go off of is their marketing material for what 10nm means.
From Intel themselves [1] TSMC and Samsung 10nm are denser than Intel 14nm. But they're closer to Intel 14nm than Intel 10nm. We'll have to see when it ships, but the expectation is that Intel 10nm will be ahead of TSMC/Samsung in density.
However, both TSMC and Samsung already ship 10nm, so it's also an important point to keep in mind. And then their 7nm processes are expected to leapfrog Intel 10nm. More details in [2], keeping in mind it's mostly announcements and estimations, so keep the salt at hand.
It's true that lithographic technologies ("10nm") don't map to a specific feature size, but TSMC 10nm is denser than Intel 14nm. You may be thinking of Intel 14nm vs TSMC 16 nm vs Samsung 14 nm, which all have comparable fin pitches, and it's certainly true that for the same technology process, Intel has been the most ambitious about what feature size that translates to in the last few years.
