Intel confirms that its first 10nm chips will roll out in H2 2017

[HEXUS]

Talk of a delay until 2018 was based upon "errors" in a job vacancy listing.

Read more.

[lumireleon]

6 Months after the intro of ZEN 14nm here comes 10nm from intel! competition is stiff.

[Nifl]

It's also worth pointing out that 14nm to 10nm seems like a small step, but it's actually huge if you look at the percentages.

[DanceswithUnix]

It's also worth pointing out that 14nm to 10nm seems like a small step, but it's actually huge if you look at the percentages.

It is a full node, so (14^2)/(10^2) says area has roughly halved (the usual definition of a full node change).

[Zimano]

Trying to remember back now, did 0.18 micron = 180nm? When do we start using the next unit down?

[DanceswithUnix]

Trying to remember back now, did 0.18 micron = 180nm? When do we start using the next unit down?

A silicon atom is about 0.2nm across, so at 1nm you are making a transistor out of a handful of atoms. I think what we call it is the last of our problems

http://www.extremetech.com/computing...on-lithography

Foundaries sound confident of getting down to 5nm, beyond there sounds interesting.

[The Hand]

This talk of 10nm next year feels pretty extreme considering AMD are still trying to flog 32nm FX chips to us... I wish AMD were releasing Zen sooner.

[scaryjim]

... I wish AMD were releasing Zen sooner.

I'd rather they released it when it's ready and got it right. AMD's been hamstrung over the last decade by a string of sub-par first gen products: first gen DX10 GPUs, first gen Phenom, Bulldozer .. all products that were significantly underbaked when they hit the shelves, and were subsequently improved upon in future iterations (I'd love to know how an 8 core Excavator FX CPU would perform, for instance).

Problem is they can't afford any more of those. They've put it right in the graphics department - I can't remember the last genuinely new AMD graphics product that underwhelmed on launch. They haven't all been world beaters, but they've all been good products. but another lacklustre CPU launch followed by a year and a bit of tweaking to catch up with Intel's previous generation mid-range platform?

I dunno; I guess we'll see before that year's out. Reading back on some of the recent AMD CPU releases has got me all maudlin and depressed: it seems we've been saying the same thing for almost 8 years now...

[CAT-THE-FIFTH]

Since Kaby Lake is being released at the end of the year,I expect it will be an Atom chip of some sort,so they can say they have 10NM chips next year for investors. I expect it will be 2018 uintil we see any 10NM desktop chips.

[Percy1983]

I really do hope Zen is done right to have a chance if intel manage this.

Personally if they take it to the i5s with better performance and cheaper then I am in, I generally struggle to justify i7 prices but if AMD can't take the top spot a good second at a good price would be great.

[Michael H]

First 10nm chip will likely be a Core M again. Relatively small die with high margins to cover lower yields, and a product most likely to benefit from the die shrink due to it's power constrained nature.

Perhaps limited to a single halo product such as a MacBook Air or Surface Pro to ensure adequate supply.