Intel refutes reports that it's ending work on 10nm, say it's making 'good progress'

Reports are starting to swirl that Intel is ending its efforts in the 10-nanometer manufacturing space, and they all point to an article posted today at SemiAccurate. According to the article, "Intel just pulled the plug on their struggling 10nm process," which it goes on to say is the "right thing to do." Whether that would be the right move or not is moot—Intel says the reports are wrong.

I reached out to Intel on the matter and was told point blank that the reports are "untrue." Intel also referred me to a Twitter post saying essentially the same thing, adding that it's "making good progress on 10nm" and that "yields are improving."

That bit about yields being "consistent with the timeline" Intel recently shared means that Cannon Lake is still on track to ship in volume during "holiday 2019." That's the last official word we've heard from Intel, anyway.

Getting to 10nm has not been without challenges. Die shrinks are tricky at this level, and Intel would (and does) argue that its manufacturing process is more advanced than the competition. Intel's Mark Bohr argued this in a blog post last year, saying that some companies have "continued to advance node names, even in cases where there was minimal or no density increase."

"The industry needs a standardized density metric to level the playing field. Customers should be able to readily compare various process offerings of a chip maker, and those of different chip makers. The challenge is in the increasing complexity of semiconductor processes, and in the variety of designs," Mohr said.

It's worth a read, as he talks about gate pitch, logic transistor density, SRAM cell size, and other technical details in what amounts to a short blog post.

At the same time, 10nm has not arrived as quickly as we had hoped, at least not in volume. The only 10nm chip from Intel is a mobile CPU without integrated graphics, and availability is comparatively limited.

It wouldn't shock us if Intel ultimately delayed its mass 10nm rollout one more time, but as it stands, Cannon Lake is still slated to arrive at the end of next year.